Signal estimation device, signal estimation method, and computer program

Abstract

To provide a signal estimation device capable of accurately estimating a transmission signal on the basis of a reception signal.SOLUTION: A signal estimation device 2 comprises: parameter setting means 213 for classifying, on the basis of system information, a plurality of reception signal components included in a reception signal y and a plurality of transmission signal components included in a transmission signal x into a plurality of signal groups X and Y depending on the reception signal components, the transmission signal components, and strength and weakness of correlation and, for each signal group, setting parameters β, ρ, τ, and η stipulating a content of a signal estimation method to any of a learning value capable of being learned by learning operation and a fixed value unrelated to the learning operation; and signal estimation means 212 for estimating, for each signal group, an estimation signal xe, which is an estimated value of the transmission signal, from the reception signal, using the signal estimation method stipulated by the set parameters.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a signal estimation device and a signal estimation method for estimating a transmission signal transmitted by a transmitting device based on a received signal received by a receiving device, and to a computer program for executing the above-mentioned signal estimation method. [Background technology]

[0002] In recent years, research has been conducted into communication systems that transmit and receive signals using Orbital Angular Momentum-Multiple Input Multiple Output (OAM-MIMO) transmission technology. The OAM-MIMO transmission technology is a transmission technology that transmits signals multiplexed using multiple radio waves having different OAM modes from a transmitting device having multiple antenna elements to a receiving device having multiple antenna elements. An example of a communication system that employs such an OAM-MIMO transmission technology is described in Non-Patent Document 1. Specifically, Non-Patent Document 1 describes a communication system that transmits signals multiplexed using multiple radio waves having different OAM modes between a transmitting device equipped with a transmitting antenna including multiple antenna elements and a receiving device equipped with a receiving antenna including multiple antenna elements.

[0003] Other prior art documents related to the present invention include Patent Documents 1 to 2 and Non-Patent Document 2. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 6583292 [Patent Document 2] Patent No. 5317021 [Non-patent literature]

[0005] [Non-Patent Document 1] Hirofumi Sasaki et al., “Experiment on Over-100-Gbps Wireless Transmission with OAM-MIMO Multiplexing System in 28-GHz Band”, 2018 IEEE Global Communications Conference (GLOBECON), December 2018 [Non-Patent Document 2] Masashi Hirabe et al., “40mTransmission of OAM mode and Polarization Multiplexing in E-band”, 2019 IEEE Global Communications Conference (GLOBECON), December 2019 Summary of the Invention [Problem to be solved by the invention]

[0006] In a communication system employing OAM-MIMO transmission technology, a receiving device receives a transmission signal transmitted by a transmitting device as a received signal and estimates the transmission signal based on the received signal. Typically, the receiving device estimates the transmission signal based on a transmission path matrix (in other words, a channel matrix) indicating the state of the transmission path between the transmitting device and the receiving device and the received signal. When the transmission signal is estimated in this way, it is desirable to estimate the transmission signal with high accuracy.

[0007] Here, when the transmitting antenna and the receiving antenna are arranged in an ideal arrangement (i.e., a plurality of antenna elements are arranged in an ideal arrangement), for example, the transmitting signal can be accurately estimated by an existing signal estimation method. An example of the existing signal estimation method is a signal estimation method based on Minimum Mean Square Error (MMSE) as shown in Equation 1. However, in reality, the transmitting antenna and the receiving antenna are not always arranged in an ideal arrangement. In this case, a general signal estimation method has a technical problem that it is not always possible to accurately estimate the transmitting signal because it does not take into account a situation in which the transmitting antenna and the receiving antenna are not arranged in an ideal arrangement.

[0008] Furthermore, even in a communication system in which a signal multiplexed using multiple radio waves having different OAM modes is transmitted from a transmitting device having a single transmitting antenna to a receiving device having a single receiving antenna, it is still desirable to accurately estimate the transmitted signal.

[0009] The present invention aims to provide a signal estimation device, a signal estimation method, and a computer program capable of solving the above-mentioned technical problems. As an example, the present invention aims to provide a signal estimation device, a signal estimation method, and a computer program capable of accurately estimating a transmitted signal based on a received signal. [Means for solving the problem]

[0010] One aspect of the signal estimation device is a signal estimation device for estimating the transmission signal from the received signal in a communication system including a transmission device that transmits a transmission signal using a transmission antenna including a plurality of transmission antenna elements, and a reception device that receives the transmission signal transmitted by the transmission device as a received signal using a reception antenna including a plurality of reception antenna elements, the signal estimation device estimating the transmission signal from the received signal, based on system information including antenna information regarding the transmission antenna and the reception antenna, distance information regarding a transmission distance between the transmission antenna and the reception antenna, carrier information regarding a carrier used by the communication system, and processing information regarding signal processing performed by the transmission device and reception processing performed by the reception device. The signal estimation device includes a parameter setting means for classifying a plurality of received signal components contained in the received signal and received by each of the plurality of receiving antenna elements, and a plurality of transmitted signal components contained in the transmitted signal and transmitted by each of the plurality of transmitting antenna elements, into a plurality of signal groups according to the strength of correlation between the received signal components and the transmitted signal components, and for setting a parameter defining the content of a signal estimation method for each of the signal groups to either a learned value that can be learned by a learning operation or a fixed value unrelated to the learning operation, and a signal estimation means for estimating an estimated signal that is an estimated value of the transmitted signal from the received signal, for each of the signal groups, using the signal estimation method defined by the parameter set by the parameter setting means.

[0011] One aspect of the signal estimation method is a signal estimation method for estimating a transmission signal from a received signal in a communication system including a transmission device that transmits a transmission signal using a transmission antenna including a plurality of transmission antenna elements, and a receiving device that receives the transmission signal transmitted by the transmission device as a received signal using a receiving antenna including a plurality of receiving antenna elements, the signal estimation method including: antenna information regarding the transmission antenna and the receiving antenna; distance information regarding a transmission distance between the transmission antenna and the receiving antenna; carrier information regarding a carrier used by the communication system; processing information regarding signal processing performed by the transmission device and reception processing performed by the receiving device; a plurality of received signal components contained in the received signal and received by each of the plurality of receiving antenna elements and a plurality of transmitted signal components contained in the transmitted signal and transmitted by each of the plurality of transmitting antenna elements are classified into a plurality of signal groups according to the strength of correlation between the received signal components and the transmitted signal components based on system information including the system information, a parameter defining the content of a signal estimation method for each of the signal groups is set to either a learned value that can be learned by a learning operation or a fixed value unrelated to the learning operation, and an estimated signal that is an estimated value of the transmitted signal from the received signal is estimated for each of the signal groups using the signal estimation method defined by the set parameter.

[0012] One aspect of the computer program is a computer program that causes a computer to execute a signal estimation method, the signal estimation method being a signal estimation method for estimating a transmission signal from a received signal in a communication system including a transmission device that transmits a transmission signal using a transmission antenna including a plurality of transmission antenna elements, and a receiving device that receives the transmission signal transmitted by the transmission device as a received signal using a receiving antenna including a plurality of receiving antenna elements, the signal estimation method including: antenna information regarding the transmission antenna and the receiving antenna; distance information regarding a transmission distance between the transmission antenna and the receiving antenna; carrier information regarding a carrier used by the communication system; and signal processing and pre-processing performed by the transmission device. Based on system information including processing information related to the reception processing performed by the receiving device, a plurality of reception signal components contained in the reception signal and received by each of the plurality of receiving antenna elements and a plurality of transmission signal components contained in the transmission signal and transmitted by each of the plurality of transmitting antenna elements are classified into a plurality of signal groups according to the strength of correlation between the reception signal components and the transmission signal components, and a parameter that specifies the content of a signal estimation method for each of the signal groups is set to either a learning value that can be learned by a learning operation or a fixed value unrelated to the learning operation, and an estimated signal that is an estimated value of the transmission signal from the reception signal for each of the signal groups is estimated using the signal estimation method specified by the set parameter. Effect of the Invention

[0013] According to each aspect of the above-mentioned signal estimation device, signal estimation method, and computer program, it is possible to accurately estimate a transmitted signal based on a received signal. [Brief description of the drawings]

[0014] [Figure 1] FIG. 1 is a block diagram showing the configuration of a communication system according to the present embodiment. [Diagram 2] FIG. 2 is a vector showing a transmission signal transmitted by a transmitting device. [Diagram 3]FIG. 3 is a vector showing a received signal received by a receiving device. [Figure 4] FIG. 4 shows the relationship between the transmitted signal and the received signal. [Diagram 5] FIG. 5 is a block diagram showing the configuration of the parameter setting unit. [Figure 6] FIG. 6 is a flowchart showing the flow of the parameter setting operation performed by the parameter setting unit. [Figure 7] FIG. 7 shows a submatrix of the transmission channel matrix. [Figure 8] FIG. 8 shows how the elements of the transmission channel matrix are rearranged. [Figure 9] FIG. 9 is a flowchart showing the flow of the signal estimation operation performed by the signal estimation unit. [Figure 10] FIG. 10 is a flowchart showing the flow of a modified example of the signal estimation operation performed by the signal estimation unit. [Figure 11] FIG. 11 is a block diagram showing a configuration of a modified example of the communication system. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] Hereinafter, with reference to the drawings, a description will be given of an embodiment of a signal estimation device, a signal estimation method, and a computer program using a communication system SYS to which the embodiment of the signal estimation device, the signal estimation method, and the computer program are applied. However, the present invention is not limited to the embodiment described below.

[0016] <1> Configuration of the communication system SYS First, the configuration of a communication system SYS of the present embodiment will be described with reference to Fig. 1. Fig. 1 is a block diagram showing the configuration of a communication system SYS of the first embodiment.

[0017] As shown in Fig. 1, the communication system SYS includes a transmitting device 1 and a receiving device 2. The transmitting device 1 transmits a transmission signal x to the receiving device 2 via a transmission path 3. The receiving device 2 receives the transmission signal x transmitted by the transmitting device 1 via the transmission path 3 as a received signal y. The transmission path 3 may include a wired transmission path (i.e., a communication network). The transmission path 3 may also include a wireless transmission path (i.e., a communication network).

[0018] The transmitting device 1 transmits a transmission signal x using an OAM (Orbital Angular Momentum) transmission technique. The OAM transmission technique is a technique for transmitting a signal multiplexed using a plurality of radio waves having different OAM modes from the transmitting device 1 to the receiving device 2. In order to transmit the transmission signal x using the OAM transmission technique, the transmitting device 1 includes a transmitting antenna 10, a signal processing device 11, and a storage device 12. The signal processing device 11 and the storage device 12 may be connected via a data bus 13.

[0019] The transmitting antenna 10 includes a plurality of array antennas 100. FIG. 1 shows an example in which the plurality of array antennas 100 are a plurality of circular array antennas (UCA: Uniform Circular Array) that are different in diameter from each other and arranged in a concentric shape. Therefore, in the example shown in FIG. 1, the plurality of array antennas 100 are arranged such that array antennas 100#1 to 100#M are arranged in this order from the center toward the outside. Each array antenna 100 includes a plurality of antenna elements 101. In the following description, it is assumed that the transmitting antenna 10 includes M (where M is a constant indicating an integer equal to or greater than 1) array antennas 100 (specifically, array antennas 100#1 to 100#M). Note that when the transmitting antenna 10 includes two or more array antennas 100 (and further when the receiving antenna 20 described later includes two or more array antennas 200), the OAM transmission technology may be referred to as an OAM-MIMO (Multiple Input Multiple Output) transmission technology. 1, each array antenna 100 includes N (where N is a constant indicating an integer of 2 or more) antenna elements 101 (specifically, antenna elements 101#1 to 101#N) arranged in a circle at equal intervals. In addition, in FIG. 1, antenna elements 101#1 to 101#N included in array antenna 100#j (j is a variable indicating an integer of 1 or more and M or less) are represented as antenna elements 101#1(#j) to 101#N(#j), respectively.

[0020] Radio waves of different OAM modes are assigned to the N antenna elements 101 included in each array antenna 100 within the same frequency band. That is, in this embodiment, the transmitting device 1 multiplexes the transmission signal x using the N OAM modes. In this case, each array antenna 100 can transmit the transmission signal x in units of N signal sets corresponding to the N OAM modes, respectively, using the N antenna elements 101. However, radio waves of the same OAM mode may be assigned to at least two antenna elements 101 included in each array antenna 100. That is, the number of OAM modes may be less than the number of antenna elements 101 included in each array antenna 100.

[0021] Furthermore, since the transmitting antenna 10 includes M array antennas 100, the N transmission signal sets can include M transmission signal components. Therefore, the transmitting device 1 can transmit a transmission signal (i.e., a transmission signal sequence) x in which a maximum of M×N transmission signal components are multiplexed. Hereinafter, a transmission signal component transmitted by the t-th (t is a variable indicating an integer not less than 1 and not more than N) antenna element 101#t (#j) of the array antenna 100#j is referred to as a "transmission signal x[j, t]". In this case, the transmission signal x can be expressed by a vector shown in FIG. 2. In FIG. 2, the "transmission signal x[j]" includes N transmission signals x[j, 1] to x[j, N] transmitted by the N antenna elements 101 included in the array antenna 100#j. In this embodiment, unless otherwise specified, the initial values ​​of the "variables (e.g., the variable t mentioned in this paragraph)" and indexes are set to 1.

[0022] The signal processing device 11 may include at least one of a central processing unit (CPU), a graphic processing unit (GPU), and a field programmable gate array (FPGA). The signal processing device 11 may read a computer program. For example, the signal processing device 11 may read a computer program stored in a storage device 12. For example, the signal processing device 11 may read a computer program stored in a computer-readable storage medium using a storage medium reading device (not shown). The signal processing device 11 may acquire (i.e., download or read) a computer program from a device (not shown) disposed outside the transmitting device 1 via a communication device (not shown). The signal processing device 11 executes the read computer program. As a result, a logical function block for executing the operation to be performed by the transmitting device 1 is realized in the signal processing device 11. Specifically, a logical function block for executing a transmission operation to transmit a transmission signal x is realized in the signal processing device 11. In other words, the signal processing device 11 can function as a controller for realizing a logical function block for executing the operation to be performed by the transmitting device 1.

[0023] Fig. 1 shows an example of logical functional blocks implemented in the signal processing device 11 to perform a transmission operation. As shown in Fig. 1, M conversion units 111 (specifically, conversion unit 111#1 to conversion unit 111#M) are implemented in the signal processing device 11.

[0024] Each transformation unit 111#j performs pre-processing (signal processing) on ​​N transmission signals x[j, 1] to x[j, N] transmitted by one array antenna 100#j corresponding to each transformation unit 111#j among the M array antennas 100. In this embodiment, each transformation unit 111#j performs inverse discrete Fourier transform (IDFT) processing on the N transmission signals x[j, 1] to x[j, N]. In this case, each transformation unit 111#j performs an N-row by N-row Fourier transform matrix F N The signal processing device 11 may include an inverse discrete Fourier transformer of length N that performs inverse discrete Fourier transform processing using the inverse discrete Fourier transform processing. The inverse discrete Fourier transform processing may be at least a part of the processing for OAM multiplexing N transmission signals x[j, 1] to x[j, N]. That is, the signal processing device 11 may divide the transmission signal x including M×N transmission signals x[1, 1] to x[M, N] into M signal series x[j] each including N transmission signals x[j, 1] to x[j, N], and perform inverse discrete Fourier transform processing on each of the M signal series x[j] to OAM multiplex the transmission signal x. The N transmission signals x[j, 1] to x[j, N] that have been subjected to the inverse discrete Fourier transform processing are transmitted to the receiving device 2 by the corresponding array antenna 100#j. In the following description, the N transmission signals x[j, 1] to x[j, N] after inverse discrete Fourier transform processing will be referred to as "transmission signals x'[j, 1] to x'[j, N]" to distinguish them from the N transmission signals x[j, 1] to x[j, N] before inverse discrete Fourier transform processing.

[0025] The storage device 12 can store desired data. For example, the storage device 12 may temporarily store a computer program executed by the signal processing device 11. The storage device 12 may temporarily store data that is temporarily used by the signal processing device 11 when the signal processing device 11 is executing a computer program. The storage device 12 may store data that the transmitting device 1 stores for a long period of time. The storage device 12 may include at least one of a RAM (Random Access Memory), a ROM (Read Only Memory), a hard disk device, a magneto-optical disk device, an SSD (Solid State Drive), and a disk array device.

[0026] The receiving device 2 includes a receiving antenna 20, a signal processing device 21, and a storage device 22. The signal processing device 21 and the storage device 22 may be connected via a data bus 23.

[0027] The receiving antenna 20 is an antenna for receiving a transmission signal x transmitted using an OAM transmission technique as a reception signal y. In this embodiment, the receiving antenna 20 includes a plurality of array antennas 200. FIG. 1 shows an example in which the plurality of array antennas 200 are a plurality of circular array antennas (UCA: Uniform Circular Array) having different diameters and arranged in a concentric shape. Therefore, in the example shown in FIG. 1, the plurality of array antennas 200 are arranged such that the array antennas 200#1 to 200#M are arranged in this order from the center to the outside. In the following description, an example in which the receiving antenna 20 includes M array antennas 200 (specifically, the array antennas 200#1 to 200#M) will be described. That is, in the following description, an example in which the number of the array antennas 200 included in the receiving antenna 20 is the same as the number of the array antennas 100 included in the transmitting antenna 10 will be described. However, the number of the array antennas 200 may be different from the number of the array antennas 100 included in the transmitting antenna 10. Each array antenna 200 includes a plurality of antenna elements 201. In the example shown in FIG. 1, each array antenna 200 includes N antenna elements 201 (specifically, antenna elements 201#1 to 201#N) arranged in a circle at equal intervals. That is, in the following description, an example will be described in which the number of antenna elements 201 included in the array antenna 200 is the same as the number of antenna elements 101 included in the array antenna 100. However, the number of antenna elements 201 included in the array antenna 200 may be different from the number of antenna elements 101 included in the array antenna 100. In FIG. 1, antenna elements 201#1 to 201#N included in the array antenna 200#i (i is a variable indicating an integer equal to or greater than 1 and equal to or less than M) are represented as antenna elements 201#1(#i) to 201#N(#i), respectively.

[0028] As described above, the transmitting device 1 transmits a transmission signal (i.e., a transmission signal sequence) x in which a maximum of M×N transmission signal components are multiplexed. Therefore, the receiving device 2 receives the transmission signal x in which a maximum of M×N transmission signal components are multiplexed as a received signal y. In this case, the received signal y is a signal in which a maximum of M×N reception signal components are multiplexed. Hereinafter, the received signal component received by the s-th (s is a variable indicating an integer not less than 1 and not more than N) antenna element 201#s(#i) of the array antenna 200#i is referred to as a "received signal y[i,s]". For example, the array antenna 200#i may receive, as received signals y[i,1] to y[i,N], N transmission signals x'[i,1] to x'[i,N] transmitted by one array antenna 100#i corresponding to the array antenna 200#i, using N antenna elements 201#1(#i) to 201#N(#i). Typically, the antenna element 201#s(#i) of the array antenna 200#i may receive the transmission signal x'[i,s] transmitted by the antenna element 101#s(#i) of the array antenna 100#i as a received signal y[i,s]. Therefore, the received signal y can be expressed by a vector shown in Fig. 3. In Fig. 3, the "received signal y[i]" includes N received signals y[i,1] to y[i,N] received by the N antenna elements 201 of the array antenna 200#i.

[0029] The signal processing device 21 may include at least one of a CPU, a GPU, and an FPGA. The signal processing device 21 reads a computer program. For example, the signal processing device 21 may read a computer program stored in the storage device 22. For example, the signal processing device 21 may read a computer program stored in a computer-readable recording medium using a recording medium reading device (not shown). The signal processing device 21 may acquire (i.e., download or read) a computer program from a device (not shown) disposed outside the receiving device 2 via a communication device (not shown). The signal processing device 21 executes the read computer program. As a result, a logical function block for executing an operation to be performed by the receiving device 2 is realized in the signal processing device 21. Specifically, a logical function block for executing a receiving operation to receive the received signal y is realized in the signal processing device 21. In other words, the signal processing device 21 can function as a controller for realizing a logical function block for executing an operation to be performed by the receiving device 2.

[0030] 1 shows an example of logical functional blocks implemented in the signal processing device 21 to perform a receiving operation. As shown in FIG 1, the signal processing device 21 implements M conversion units 211 (specifically, conversion units 211#1 to 211#M), a signal estimation unit 212, and a parameter setting unit 213.

[0031] Each transformation unit 211#i performs pre-processing (signal processing) on ​​N received signals y[i, 1] to y[i, N] received by one array antenna 200#i corresponding to each transformation unit 211#i among the M array antennas 200. In this embodiment, each transformation unit 211#i performs discrete Fourier transform processing (DFT) on the N received signals y[i, 1] to y[i, N]. In this case, each transformation unit 211#i performs a Fourier transform matrix F NThe signal processing device 21 may include a discrete Fourier transformer of length N that performs discrete Fourier transform processing using the above. The discrete Fourier transform processing may constitute at least a part of the processing for separating the OAM-multiplexed N received signals y[i, 1] to y[i, N]. That is, the signal processing device 21 divides the received signal y including M×N received signals y[1, 1] to y[M, N] into M signal series each including N received signals y[i, 1] to y[i, N], and performs discrete Fourier transform processing on each of the M signal series. In other words, the signal processing device 21 performs discrete Fourier transform processing on the received signal y including M×N received signals y[1, 1] to y[i, N] in units of M signal series each including N received signals y[i, 1] to y[i, N]. The received signal y that has been subjected to Fourier transform processing (i.e., the received signal y including M×N received signals y[1, 1] to y[M, N] that have been subjected to Fourier transform processing) is output to the signal estimation unit 212. In the following description, unless otherwise specified, "received signal y (including received signal y[i, s] and received signal y[i])" refers to the received signal that has been subjected to Fourier transform processing.

[0032] The signal estimation unit 212 estimates the transmission signal x based on the received signal y including the received signal y[M, N] from the M×N received signals y[1, 1] output from the M transform units 211. Specifically, the signal estimation unit 212 estimates an estimated signal x_estimate[M, N] from M×N estimated signals x_estimate[1, 1] which are estimates of the transmission signal x[M, N] from the M×N transmitted signals x[1, 1] included in the transmitted signal x based on the received signal y. That is, the signal estimation unit 212 estimates an estimated signal x_estimate including the estimated signal x_estimate[M, N] from the M×N estimated signals x_estimate[1, 1] based on the received signal y. The signal estimation unit 212 estimates an estimated signal x_estimate including the estimated signal x_estimate[j] which is an estimate of the transmission signal x[j] included in the transmitted signal x based on the received signal y. The signal estimation unit 212 estimates an estimate signal x_estimate including N estimate signals x_estimate[j,1] to x_estimate[j,N] that are estimates of N transmit signals x[j,1] to x[j,N] included in the transmit signal x[j] based on the received signal y. Note that, hereinafter, for the sake of simplicity, the "estimate signal x_estimate" will be written as the "estimate signal xe".

[0033] Here, the relationship between the transmission signal x and the reception signal y is as shown in FIG. 4. As shown in FIG. 4, the reception signal y theoretically (in other words, ideally) coincides with a signal obtained by adding noise n to the product of the transmission signal x and the transmission channel matrix H. The transmission channel matrix H is a channel matrix indicating the state (typically, the transmission state) of the transmission channel 3 between the transmission antenna 10 and the reception antenna 20. The transmission channel matrix H is a block matrix of MN rows by MN columns having a submatrix H[i, j] of N rows by N rows as an (i, j) component. The submatrix H[i, j] indicates the state of the transmission channel 3 between the array antenna 100#j of the transmission antenna 10 and the array antenna 200#i of the reception antenna 20. Moreover, "n[j]" in FIG. 4 indicates a vector having a component of noise (e.g., white noise) occurring in the signal received by the array antenna 200#j in the transmission channel 3.

[0034] Another example of a signal estimation method for estimating a transmitted signal x based on a received signal y is a signal estimation method based on a minimum mean square error (MMSE). The signal estimation method based on the minimum mean square error is a signal estimation method that estimates a transmitted signal x using Equation 1. In addition, "H†" in Equation 1 indicates a conjugate transpose matrix of a transmission channel matrix H. "σ[i]" in Equation 1 indicates a vector including N parameters "σ[i,1], σ[i,2]..., σ[i,N]" that respectively indicate the signal to noise ratios (SNRs) of the N received signals y[i,1] to y[i,N] included in the received signal y[i]. In formula 1, “D(σ[1], σ[2] , σ[N])” denotes a diagonal matrix with MN rows and MN columns, whose diagonal elements are the M × N parameters “σ[1], σ[2] , σ[N]” representing the signal-to-noise power ratios.

[0035]

number

[0036] Here, when the transmitting antenna 10 and the receiving antenna 20 are ideally arranged and the multiple antenna elements 101 and the multiple antenna elements 201 are ideally arranged, the transmitting signal x can be accurately estimated from the receiving signal y by a signal estimation method based on the minimum mean square error. Note that an example of the ideal arrangement is an arrangement in which the transmitting antenna 10 and the receiving antenna 20 are axially symmetric (i.e., coaxial).

[0037] However, in reality, the transmitting antenna 10 and the receiving antenna 20 are not always arranged in an ideal arrangement. For example, the transmitting antenna 10 and the receiving antenna 20 may not always be axially symmetrical due to at least a partial shift in the arrangement position of the transmitting antenna 10 and the receiving antenna 20 caused by wind and rain. In this case, a general signal estimation method does not take into account a situation in which the transmitting antenna 10 and the receiving antenna 20 are not always arranged in an ideal arrangement, and therefore may not be able to accurately estimate the transmission signal x.

[0038] Therefore, in this embodiment, the signal estimation unit 212 employs a new signal estimation method proposed below to estimate the transmission signal x. Specifically, the signal estimation unit 212 estimates the transmission signal x using a signal estimation method whose content is defined by the parameters set by the parameter setting unit 213 (in other words, whose content is changeable). In particular, in the following description, the signal estimation unit 212 estimates the transmission signal x using a signal estimation method whose content is defined by an arithmetic expression (i.e., a mathematical expression) including the parameters set by the parameter setting unit 213 as coefficients. The parameters set by the parameter setting unit 213 are learned by a learning operation as necessary, as will be described later in detail. As a result, the signal estimation unit 212 can estimate the transmission signal x with higher accuracy than a signal estimation method based on the minimum mean square error, even when the transmitting antenna 10 and the receiving antenna 20 are not always arranged in an ideal arrangement mode.

[0039] The parameter setting unit 213 sets parameters that define the content of the signal estimation method used by the signal estimation unit 212 to estimate the estimated signal xe (the content of the arithmetic expression in this embodiment). In this embodiment, an example in which the parameter setting unit 213 sets four types of parameters will be described. That is, in this embodiment, an example in which the content of the signal estimation method used by the signal estimation unit 212 to estimate the estimated signal xe is defined by four types of parameters will be described. In other words, in this embodiment, an example in which the content of the signal estimation method used by the signal estimation unit 212 to estimate the estimated signal xe can be changed by four types of parameters will be described. Specifically, the parameter setting unit 213 sets a parameter β, a parameter ρ, a parameter τ, and a parameter η as the four types of parameters. The parameter β is set by N×R (R is a constant indicating an integer equal to or greater than 1) parameters β1 (1) From β R (N) In the following description, N × R parameters β1 (1) From β R (N) Using the variables k and r as appropriate, we define the parameter β r (k) Each parameter β r (k) is, for example, a vector including M-dimensional vector components. The variable k indicates an integer greater than or equal to 1 and less than or equal to N. The variable r indicates an integer greater than or equal to 1 and less than or equal to R. In addition, the parameter ρ is an N×R number of parameters ρ1 (1) From ρ R (N) In the following description, N × R parameters ρ1 (1) From ρ R (N) Using the variables k and r as appropriate, we define the parameter ρ r (k) Each parameter ρ r (k) is, for example, a vector containing M-dimensional vector components. Also, the parameter τ is a set of N × R parameters τ1 (1) From τ R (N)In the following description, N × R parameters τ1 (1) From τ R (N) Using the variables k and r as appropriate, we define the parameter τ r (k) Each parameter τ r (k) is, for example, a vector containing M-dimensional vector components. The parameter η is a set of N×R parameters η1 (1) From R (N) In the following description, N × R parameters η1 (1) From R (N) Using the variables k and r as appropriate, we define the parameter η r (k) Each parameter η r (k) is, for example, a vector containing M-dimensional vector components.

[0040] The configuration of such a parameter setting unit 213 is shown in Fig. 5. As shown in Fig. 5, the parameter setting unit 213 includes a transmission channel model generating unit 2131, a signal classifying unit 2132, a communication capacity calculating unit 2133, and a parameter setting unit 2134. Note that the operations of the transmission channel model generating unit 2131, the signal classifying unit 2132, the communication capacity calculating unit 2133, and the parameter setting unit 2134 of the parameter setting unit 213 will be described in detail later, and therefore will not be described here.

[0041] The storage device 22 can store desired data. For example, the storage device 22 may temporarily store a computer program executed by the signal processing device 21. The storage device 22 may temporarily store data that is temporarily used by the signal processing device 21 when the signal processing device 21 is executing a computer program. The storage device 22 may store data that the receiving device 2 stores for a long period of time. The storage device 22 may include at least one of a RAM, a ROM, a hard disk device, a magneto-optical disk device, an SSD, and a disk array device.

[0042] <2> Operation of the communication system SYS Next, the operation of the communication system SYS will be described. As described above, the communication system SYS mainly uses the signal estimation unit 212 to perform a signal estimation operation for estimating an estimated signal xe from a received signal y. Furthermore, the communication system SYS mainly uses the parameter setting unit 213 to perform a parameter setting operation for setting parameters β, ρ, τ, and η that define the contents of a signal estimation method used by the signal estimation unit 212 to estimate an estimated signal xe from a received signal y. Therefore, the parameter setting operation and the signal estimation operation will be described in order below.

[0043] <3-1> Parameter setting operation performed by the parameter setting unit 213 First, the parameter setting operation performed by the parameter setting unit 213 will be described with reference to Fig. 6. Fig. 6 is a flowchart showing the flow of the parameter setting operation performed by the parameter setting unit 213.

[0044] As shown in FIG. 6, first, the transmission path model generating unit 2131 acquires system information (step S10). The system information includes basic information of the communication system SYS. For example, the system information may include antenna information related to the transmitting antenna 10 and the receiving antenna 20. The antenna information may include information related to the number of array antennas 100 (the above-mentioned constant M in this embodiment) included in the transmitting antenna 10 and the number of antenna elements 101 (the above-mentioned constant N in this embodiment) included in the array antenna 100. The antenna information may include information related to the number of array antennas 200 (the above-mentioned constant M in this embodiment) included in the receiving antenna 20 and the number of antenna elements 201 (the above-mentioned constant N in this embodiment) included in the array antenna 200. The system information may include distance information related to the transmission distance between the transmitting antenna 10 and the receiving antenna 20 (i.e., the distance of the transmission path 3). The distance information may include information related to the distances of MN×MN transmission paths secured between the M×N antenna elements 101 included in the transmitting antenna 10 and the M×N antenna elements 201 included in the receiving antenna 20. When the transmission distance between the transmitting antenna 10 and the receiving antenna 20 is uniquely determined, the relative positional relationship (i.e., arrangement) between the transmitting antenna 10 and the receiving antenna 20 is uniformly determined. Therefore, the distance information may be considered to be information related to the relative positional relationship (i.e., arrangement) between the transmitting antenna 10 and the receiving antenna 20. The system information may include carrier information related to the carrier used by the communication system SYS. The carrier information may include information related to at least one of the frequency and wavelength of the carrier. The system information may include processing information related to the signal processing performed by the transmitting device 1 and the receiving processing performed by the receiving device. As described above, the transmitting device 1 performs inverse discrete Fourier transform processing. Therefore, the processing information may include information related to the inverse discrete Fourier transform processing. The information related to the inverse discrete Fourier transform processing is calculated based on the above-mentioned N-row x N-column Fourier transform matrix F NAs described above, the receiving device 2 performs discrete Fourier transform processing. Therefore, the processing information may include information about the discrete Fourier transform processing. The information about the discrete Fourier transform processing may include the above-mentioned N-row × N-column Fourier transform matrix F N It may also include information regarding.

[0045] Thereafter, the transmission channel model generating unit 2131 generates a transmission channel matrix H based on the system information acquired in step S10 (step S11). The transmission channel model generating unit 2131 generates a transmission channel matrix H indicating a theoretical state of the transmission channel 3 derived from the system information, instead of a transmission channel matrix H indicating an actual state of the transmission channel 3 (typically, a transmission channel matrix H generated based on an actual measurement value of the state of the transmission channel 3). For this reason, the transmission channel model generating unit 2131 may be considered to calculate a theoretical value, an estimated value, or an approximate value of the transmission channel matrix H.

[0046] As shown in Fig. 4, the transmission path matrix H has a submatrix H[i,j] as an (i,j) component, which indicates the state of the transmission path 3 between the array antenna 100#j of the transmitting antenna 10 and the array antenna 200#i of the receiving antenna 20. As shown in Fig. 7, the submatrix H[i,j] is a Fourier transform matrix F N and component h[i, j] s、t As the sth row and tth column components, an N row × N column matrix is ​​given by N The matrix h[i, j] corresponds to the matrix obtained by multiplying the inverse matrix of s、t corresponds to the radio wave propagation gain between the array antenna 100#j and the array antenna 200#i (more specifically, the radio wave propagation gain between the antenna element 101#t(#j) and the antenna element 201#s(#i)). Therefore, the component in the sth row and tth column of the submatrix H[i,j] is the component h[i,j] s、t Typically, the component h[i,j] indicates the state of the transmission path 3 between the antenna element 101#t(#j) of the array antenna 100#j and the antenna element 201#s(#i) of the array antenna 200#i. s、tcan be calculated using Equation 2. In Equation 2, G is a constant that is determined in advance depending on factors other than radio wave propagation. In Equation 2, d(i, j; s, t) indicates the transmission distance between the t-th antenna element 101#t(#j) of the array antenna 100#j and the s-th antenna element 201#s(#i) of the array antenna 200#i. In Equation 2, λ indicates the wavelength of the carrier wave used by the communication system SYS. In Equation 2, "exp" indicates an exponential function with Napier's number as the base.

[0047]

number

[0048] For this reason, in order to generate the transmission path matrix H, the transmission path model generating unit 2131 uses distance information (i.e., transmission distance d(i, j; s, t)) indicated by the system information, carrier information (i.e., wavelength λ) indicated by the system information, and Equation 2 to calculate N×N components h[i, j] s、t After that, the transmission channel model generating unit 2131 calculates the processing information indicated by the system information (i.e., the Fourier transform matrix F N ) and the calculated N × N components h[i, j] s、t and generate a submatrix [i, j]. Thereafter, the transmission channel model generation unit 2131 generates submatrices H[1, 1] to H[M, M] by repeating the same operation while changing the variable i between 1 and M and changing the variable j between 1 and M. As a result, the transmission channel matrix H is generated. At this time, the transmission channel model generation unit 2131 may determine the number of rows and the number of columns of each of the submatrix H[i, j] and the transmission channel matrix H based on the antenna information (i.e., the numbers of the array antennas 100 and 200 and the antenna elements 101 and 201) indicated by the system information.

[0049] 6 again, thereafter, the signal classification unit 2132 classifies the M×N transmission signals x[1, 1] to x[M, N] included in the transmission signal x into a plurality of signal groups X (i.e., a group considered to be composed of at least two of the transmission signals x[1, 1] to x[M, N]) based on the transmission channel matrix H generated in step S11 (step S12). In this embodiment, the signal classification unit 2132 classifies the M×N transmission signals x[1, 1] to x[M, N] into N signal groups X (1) From X (N) Based on the transmission channel matrix H generated in step S11, the signal classification unit 2132 also classifies the M×N received signals y[1, 1] to y[M, N] included in the received signal y into a plurality of signal groups Y (i.e., groups considered to be composed of at least two of the received signals y[1, 1] to y[M, N]) (step S12). In this embodiment, the signal classification unit 2132 classifies the M×N received signals y[1, 1] to y[M, N] into N signal groups Y (1) From Y (N) Classify into:

[0050] The signal classification unit 2132 classifies the signal group Y (k) and signal group X (k) The correlation between the received signal components and the transmitted signal components classified into signal group Y (k) Received signal components classified into signal group X (1) From X (k-1) and X (k+1) From X (N) In other words, the signal classification unit 2132 classifies the transmission signal x and the reception signal y into a plurality of signal groups according to the strength of the correlation between the reception signal y and the transmission signal x (i.e., the correlation between the M×N reception signals y[1, 1] to y[M, N] included in the reception signal y and the M×N transmission signals x[1, 1] to x[M, N] included in the transmission signal x).

[0051] In this embodiment, the state in which "the correlation between the received signal component and the first transmission signal component is stronger than the correlation between the received signal component and the second transmission signal component" may include a state in which the influence of the first transmission signal component on the received signal component is greater than the influence of the second transmission signal component on the received signal component. The state in which "the correlation between the received signal component and the first transmission signal component is stronger than the correlation between the received signal component and the second transmission signal component" may include a state in which the amount of change in the received signal component when the first transmission signal component changes by a predetermined amount is greater than the amount of change in the received signal component when the second transmission signal component changes by the same predetermined amount.

[0052] In this embodiment, the signal classification unit 2132 uses the magnitude of the absolute value of each component of the transmission path matrix H as the strength of correlation between one transmission signal component and one reception signal component corresponding to each component. Specifically, the component in the sth row and tth column of the submatrix H[i, j] constituting the transmission path matrix H indicates the state of the transmission path 3 between the tth antenna element 101#t(#j) of the array antenna 100#j of the transmission antenna 10 and the sth antenna element 201#s(#i) of the array antenna 200#i of the reception antenna 20. Therefore, the magnitude of the absolute value of the component in the sth row and tth column of the submatrix H[i, j] indicates the strength of correlation between the transmission signal x[j, t] transmitted by the tth antenna element 101#t(#j) of the array antenna 100#j and the reception signal y[i, s] received by the sth antenna element 201#s(#i) of the array antenna 200#i. Therefore, the signal classification unit 2132 classifies the transmission signal components and the reception signal components corresponding to the components of the transmission channel matrix H having a relatively large absolute value into signal groups having the same index (i.e., signal group X (k) and signal group Y (k) ) are classified into the transmitted signal x and the received signal y.

[0053] In order to classify the transmission signal x and the reception signal y, the signal classification unit 2132 rearranges the order of the MN×MN elements included in the transmission channel matrix H generated in step S11. Specifically, the signal classification unit 2132 rearranges the order of the MN×MN elements included in the transmission channel matrix H so that the transmission channel matrix H becomes a block diagonal matrix as shown in Fig. 8. The block diagonal matrix may be a matrix in which the absolute values ​​of the elements other than the diagonal block elements are smaller than the absolute values ​​of the diagonal block elements. Typically, the block diagonal matrix may be a matrix in which the elements other than the diagonal block elements are zero. Specifically, as shown in the lower part of Fig. 8, the transmission channel matrix H in which the order of the MN×MN elements has been rearranged is a submatrix H (k、m) is regarded as a block matrix with MN rows and MN columns having (k, m) elements (where m is a variable indicating an integer between 1 and N). When k=m is satisfied, the transmission channel matrix H is expressed as a submatrix H (k、m) If k ≠ m is satisfied while not being a zero matrix, then the submatrix H (k、m) becomes a zero matrix. Alternatively, the transmission channel matrix H is a submatrix H when k=m is satisfied. (k、m) The norm of (i.e., the sum of the absolute values ​​of the elements) of the submatrix H for which k ≠ m is satisfied. (k、m) The bottom part of Figure 8 shows the case where k ≠ m is satisfied, and the submatrix H (k、m) This shows an example where k is a zero matrix, but even if k ≠ m is satisfied, the submatrix H (k、m) does not have to be a zero matrix. In the following description, unless otherwise specified, the transmission channel matrix H refers to a matrix in which the order of elements is rearranged so that it becomes a block diagonal matrix.

[0054] The signal classification unit 2132 further changes the order of the M×N transmission signals x[1,1] to x[M,N] and the M×N reception signals y[1,1] to y[M,N] in the matrix operation formula (see the upper part of FIG. 8) showing the relationship between the transmission signal x and the reception signal y in accordance with the change in the order of the components of the transmission channel matrix H. Specifically, the signal classification unit 2132 changes the order of the M×N transmission signals x[1,1] to x[M,N] and the order of the M×N reception signals y[1,1] to y[M,N] so that the relationship between the M×N transmission signals x[1,1] to x[M,N] and the M×N reception signals y[1,1] to y[M,N] does not change due to the change in the order of the components of the transmission channel matrix H. As a result, the relationship between the transmitted signal x and the received signal y changes from the relationship shown in FIG. 4 (i.e., the relationship before the order of the MN×MN components is rearranged) to the relationship shown in the lower part of FIG. 8 (i.e., the relationship after the order of the MN×MN components is rearranged).

[0055] Thereafter, the signal classification unit 2132 classifies the transmission signal x in which the order of the transmission signal components has been rearranged into N transmission signals x each including M consecutive transmission signal components. (1) From x (N) Similarly, the signal classification unit 2132 divides the received signal y with the order of the received signal components rearranged into N received signals y (1) From y (N) Here, since the transmission channel matrix H is a block diagonal matrix, the submatrix H (k、k) The norm of the submatrix H (k、1) From H (k、k-1) and H (k、k+1) From H (k、N) Therefore, the transmission channel matrix H is larger than the norm of the received signal y (k) and the transmitted signal x (k) The correlation with the received signal y (k) and the transmitted signal component x (1) x (k-1) and x (k+1) From x (N) Therefore, the signal classification unit 2132 classifies the transmitted signal x (k) and the received signal y (k), respectively, for the signal group X (k) and Y (k) Classify into:

[0056] The signal classification unit 2132 outputs indices for identifying transmitted signal components classified into each signal group X (for example, the above-mentioned variable j for identifying the array antenna 100 and variable t for identifying the antenna element 101) to each of the signal estimation unit 212 and the communication capacity calculation unit 2133. The signal classification unit 2132 outputs indices for identifying received signal components classified into each signal group Y (for example, the above-mentioned variable i for identifying the array antenna 200 and variable s for identifying the antenna element 201) to each of the signal estimation unit 212 and the communication capacity calculation unit 2133.

[0057] 6 again, thereafter, the communication capacity calculation unit 2133 calculates a communication capacity C indicating an amount of information (particularly, its maximum value) that can be transmitted via the transmission path 3 based on the transmission path matrix H and the classification result of the signal groups X and Y by the signal classification unit 2132 (step S13). In this embodiment, the communication capacity calculation unit 2133 calculates three types of communication capacities C1, C2, and C3.

[0058] The communication capacity C1 indicates the amount of information that can be transmitted when the transmission signal x is transmitted using all of the MN×MN transmission paths secured between the M×N antenna elements 101 of the transmitting antenna 10 and the M×N antenna elements 201 of the receiving antenna 20. In this case, the estimated value of the transmission signal component transmitted by each antenna element 101 is estimated using all of the M×N received signals y[1, 1] to y[M, N] received by the M×N antenna elements 201, respectively. For this reason, the communication capacity C1 may be considered to indicate the amount of information that can be transmitted when the estimated value of the transmission signal component transmitted by each antenna element 101 is estimated using all of the M×N received signals y[1, 1] to y[M, N]. In other words, the communication capacity C1 may be regarded as indicating the amount of information that can be transmitted when the estimated signal xe[t, j], which is an estimate of the transmitted signal x[j, t] transmitted by the t-th antenna element 101#t(#j) of the array antenna 100#j, is estimated using all of the M×N received signals y[1, 1] to y[M, N].

[0059] The communication capacity calculation unit 2133 may calculate the communication capacity C1 by using Equation 3. In Equation 3, Λ(1) to Λ(MN) respectively indicate M×N eigenvalues ​​of the product matrix of the transmission channel matrix H and the conjugate transpose matrix H†. In Equation 3, SNR indicates a signal power to noise power ratio.

[0060]

number

[0061] The communication capacity C2 indicates the amount of information that can be transmitted when transmitting the transmission signal x using a combination of the antenna element 101 and the antenna element 201 corresponding to a combination of a transmission signal component and a reception signal component that are classified into the signal groups X and Y having the same index (i.e., relatively strong correlation). In particular, the communication capacity C2 indicates a theoretical value (e.g., theoretical limit value) of the amount of information that can be transmitted when transmitting the transmission signal x using a combination of the antenna element 101 and the antenna element 201 corresponding to a combination of a transmission signal component and a reception signal component that are relatively strong correlation. In other words, the communication capacity C2 indicates a theoretical value of the amount of information that can be transmitted when transmitting the transmission signal x without using a combination of the antenna element 101 and the antenna element 201 corresponding to a combination of a transmission signal component and a reception signal component that are classified into the signal groups X and Y having different indexes (i.e., relatively weak correlation). In this case, the estimated value of the transmission signal component transmitted by each antenna element 101 is estimated from the reception signal component that is classified into the signal group Y having the same index as the signal group X to which the transmission signal component is classified. That is, the estimated value of the transmission signal component transmitted by each antenna element 101 is estimated from the received signal component having a relatively strong correlation with the transmission signal component. On the other hand, the estimated value of the transmission signal component transmitted by each antenna element 101 is estimated without using the received signal component having a relatively weak correlation with the transmission signal component. For this reason, the communication capacity C2 may be regarded as indicating a theoretical value of the amount of information that can be transmitted when the estimated value of the transmission signal component transmitted by each antenna element 101 is estimated using the received signal component having a relatively strong correlation with the transmission signal component but without using the received signal component having a relatively weak correlation with the transmission signal component.

[0062] The communication capacity calculation unit 2133 may calculate the communication capacity C2 by using Equation 4. In Equation 4, λ(k,1) to λ(k,M) are the submatrix H (k、k) and the corresponding submatrix H (k、k) The matrix H is the conjugate transpose of (k、k) † respectively denote the M eigenvalues ​​of the product matrix with . In Equation 4, SNR denotes the signal power to noise power ratio.

[0063]

number

[0064] Like the communication capacity C2, the communication capacity C3 indicates the amount of information that can be transmitted when the transmission signal x is transmitted using a combination of the antenna element 101 and the antenna element 201 corresponding to a combination of a transmission signal component and a reception signal component that have a relatively strong correlation. The communication capacity C3 differs from the communication capacity C2 in that it indicates the amount of information that can be transmitted when the transmission signal x is estimated using a signal estimation method based on the minimum mean square error (or another signal estimation method different from the signal estimation method used by the signal estimation unit 212, the same applies below). In other words, the communication capacity C3 may be considered to indicate the amount of information that can be transmitted when an estimated value of a transmission signal component transmitted by each antenna element 101 is estimated using a reception signal component that has a relatively strong correlation with the transmission signal component, using a signal estimation method based on the minimum mean square error.

[0065] The communication capacity calculation unit 2133 may calculate the communication capacity C3 by using Equation 5. In Equation 5, w(k,1) to w(k,M) respectively indicate M diagonal elements of the M row by M column matrix B calculated by Equation 6. In Equation 6, I indicates an M row by M column unit matrix.

[0066]

number

[0067]

number

[0068] Thereafter, the parameter setting unit 2134 sets the parameters β, ρ, τ, and η based on the communication capacities C1 to C3 calculated in step S13 (steps S14 to S19).

[0069] In order to set the parameters β, ρ, τ, and η, the parameter setting unit 2134 judges whether the communication capacity C1 is equal to the communication capacity C2 (step S14). If it is judged that the communication capacity C1 is equal to the communication capacity C2 (step S14: Yes), the parameter setting unit 2134 sets the parameter β to a fixed value that is determined independently of the learning operation described later (step S15). Specifically, the parameter setting unit 2134 sets the parameter β to a fixed value that is determined independently of the learning operation described later. r (k) (That is, N × R parameters β1 (1) From β R (N) That is, the parameter setting unit 2134 sets the parameters β r (k) The vector components of the parameter β are set to a fixed value of zero. On the other hand, if it is determined in step S14 that the communication capacities C1 and C2 are not equal (step S14: No), the parameter setting unit 2134 sets the parameter β to a learned value that can be learned by a learning operation described later (step S16). Specifically, the parameter setting unit 2134 sets the parameter β r (k) Set the parameter β to an M-dimensional vector containing M vector components corresponding to the learnable learning values. r (k) When a value to be set as a learning value for parameter β has already been learned, the parameter setting unit 2134 may set the parameter β to a value that has already been learned as a learning value. When a value to be set as a learning value for parameter β has not yet been learned, the parameter setting unit 2134 may set the parameter β to an initial value of the learning value. The initial value may be any value. An example of the initial value is an M-dimensional vector including M values ​​whose average is 1 and whose distribution is normal.

[0070] In parallel with or before or after the operations of steps S14 to S16, the parameter setting unit 2134 judges whether the communication capacity C2 is equal to the communication capacity C3 (step S17). If it is judged in step S17 that the communication capacity C2 is equal to the communication capacity C3 (step S17: Yes), the parameter setting unit 2134 sets the parameter ρ to a fixed value (step S18). Specifically, the parameter setting unit 2134 sets the parameter ρ r (k) (That is, N × R parameters ρ1 (1) From ρ R (N) (each of which) is set to a vector including M vector components that are predetermined as fixed values ​​that do not vary depending on the variables k and r. That is, the parameter setting unit 2134 sets the parameter ρ r (k) The vector components of the parameter τ are set to fixed values. The parameter setting unit 2134 also sets the parameter τ to a fixed value (step S18). Specifically, the parameter setting unit 2134 sets the parameter τ r (k) (That is, N × R parameters τ1 (1) From τ R (N) ) to a vector including M vector components that correspond to the upper limit value (for example, a value corresponding to infinity) that can be set in the specifications of the communication system SYS. That is, the parameter setting unit 2134 sets the parameter τ r (k) The vector components of the parameter η are set to the upper limit value that can be set in the specifications of the communication system SYS. The parameter setting unit 2134 also sets the parameter η to a fixed value (step S18). Specifically, the parameter setting unit 2134 sets the parameter η r (k) (That is, N × R parameters η1 (1) From R (N) That is, the parameter setting unit 2134 sets the parameters η r (k)The vector components of are set to a fixed value of zero. On the other hand, if it is determined in step S17 that the communication capacities C2 and C3 are not equal (step S17: No), the parameter setting unit 2134 sets each of the parameters ρ, τ, and η to a learning value that can be learned by a learning operation (step S19). Note that the content of the operation of setting the learning values ​​to each of the parameters ρ, τ, and η may be the same as the content of the operation of setting the learning value to the parameter β.

[0071] Through the above operations, the parameters β, ρ, τ, and η are set, that is, the details of the signal estimation method used by the signal estimation unit 212 to estimate the estimated signal xe are set.

[0072] The learning values ​​set for each of the parameters β, ρ, τ, and η are learned by a learning operation. The learning operation may be performed by the receiving device 2, or may be performed by a device different from the receiving device 2. The receiving device 2 (or a device different from the receiving device 2, which will be the same hereinafter in the description of the learning operation) may perform the learning operation before the signal estimation unit 212 starts the signal estimation operation. The receiving device 2 may perform the learning operation online after the signal estimation unit 212 starts the signal estimation operation.

[0073] The receiving device 2 may learn the parameters β, ρ, τ, and η using a loss function Loss based on an error (e.g., square error) between an estimated signal xe (i.e., a definite estimated signal xef described in detail later) finally output by the signal estimation unit 212 and a transmission signal x actually transmitted by the transmitting device 1. Specifically, the receiving device 2 acquires learning data including a sample value of the transmission signal x (so-called a correct answer label) and a sample value of a received signal y received by the receiving device 2 when the sample value of the transmission signal x is transmitted from the transmitting device 1 to the receiving device 2. Thereafter, the receiving device 2 may learn the parameters β, ρ, τ, and η using a loss function Loss based on an error between a definite estimated signal xef output from the signal estimation unit 212 to which the sample value of the received signal y is input and the sample value of the transmission signal x.

[0074] The receiving device 2 may learn the parameters β, ρ, τ, and η using the backpropagation method used to learn the parameters of the neural network. Equation 7 shows an example of a loss function Loss based on the squared error between the deterministic estimation signal xef and the transmission signal x for learning the parameters β, ρ, τ, and η using the backpropagation method. Note that "D" in Equation 7 indicates a set of transmission signals x used to learn the parameters β, ρ, τ, and η (i.e., a set of teacher data).

[0075]

number

[0076] The effect of the learning object (for example, at least one of the parameters β, ρ, τ, and η) on the loss function Loss calculated by Equation 7 can be expressed by a partial differential that can be calculated by the backpropagation method. Therefore, the receiving device 2 may learn the learning object using Equation 8. Note that Equation 8 is used when the learning object has a parameter ρ r (k) In the formula 8, "ε" represents a hyperparameter that can be set arbitrarily.

[0077]

number

[0078] <3-2> Signal estimation operation performed by the signal estimation unit 212 Next, the signal estimation operation performed by the signal estimation unit 212 will be described with reference to Fig. 9. Fig. 9 is a flowchart showing the flow of the signal estimation operation performed by the signal estimation unit 212.

[0079] 9, first, a received signal y is input to the signal estimation unit 212 (step S20). In this embodiment, the received signal y is estimated by dividing the M×N received signal components y[1, 1] to y[M, N] contained in the received signal y into a signal group Y (1) From Y (N)That is, the received signal y is input to the signal estimation unit 212 in a state where it is classified based on the signal group Y (1) The received signal y is classified into (1) From signal group Y (N) The received signal y is classified into (N) Specifically, the signal estimation unit 212 receives the received signal y (k) is input. Note that [i, s] ∈ Y in Equation 9. (k) is the signal group Y (k) , and indicates an index for identifying the received signal components classified into the above (for example, the variable i specifying the array antenna 200 and the variable s specifying the antenna element 201 described above).

[0080]

number

[0081] However, the received signal y includes M×N received signal components y[1, 1] to y[M, N] that belong to the signal group Y (1) From Y (N) In this case, the signal estimation unit 212 may input the signal group Y (1) From Y (N) Based on the received signal y, (1) From y (N) may be classified as follows.

[0082] In parallel with or before or after the operation of step S20, the signal estimation unit 212 receives a transmission channel matrix H as initial information (step S21). The transmission channel matrix H input to the signal estimation unit 212 to perform the signal estimation operation is a transmission channel matrix generated based on the actual measurement result of the state of the transmission channel 3. In other words, the transmission channel matrix H input to the signal estimation unit 212 is different from the transmission channel matrix H generated by the parameter setting unit 213 in the parameter setting operation (that is, the theoretical value, estimated value, or approximate value of the transmission channel matrix H).

[0083] In this embodiment, the transmission channel matrix H is a signal group X (1) From X (N) and signal group Y (1) From Y (N) The elements are rearranged to form a block diagonal matrix based on the above, and the matrix is ​​input to the signal estimation unit 212. That is, the block diagonal matrix (i.e., the submatrix H (k、m) The transmission channel matrix H, which is a block matrix with MN rows and MN columns having k, m elements, is input to the signal estimation unit 212. (k、m) is the signal group Y (k) The received signal y is classified into (k) M antenna elements 201 receiving a signal group X (m) Transmitted signal x classified into (m) Specifically, the signal estimation unit 212 includes a submatrix H (k、m) may be input as a transmission channel matrix H including as a (k, m) component. (k) is signal group X (k) , and indicates an index for identifying the transmission signal components classified into (for example, the variable j specifying the array antenna 100 and the variable t specifying the antenna element 101 described above).

[0084]

number

[0085] However, the transmission channel matrix H may be input to the signal estimation unit 212 without rearranging the elements to form a block diagonal matrix. In this case, the signal estimation unit 212 estimates the signal group X (1) From X (N) and signal group Y (1) From Y (N) Based on this, the elements of the transmission channel matrix H may be rearranged so that the transmission channel matrix H becomes a block diagonal matrix.

[0086] In parallel with or before or after the operations of steps S20 to S21, the signal estimation unit 212 performs an initialization operation (step S22). Specifically, the signal estimation unit 212 sets a variable r to 1. The variable r is a value that the signal estimation unit 212 estimates based on the transmitted signal x (k) The estimated signal xe is an estimate of (k) This is because, as will be described later, the signal estimation unit 212 counts the number of times the operation of estimating the estimated signal xe (k) Therefore, in the following, the signal estimation unit 212 will refer to the estimated signal xe (k) "Estimated signal xe r+1 (k) The signal estimation unit 212 also estimates the estimated signal xe (k) The internal data v used by the signal estimation unit 212 to estimate (k) and z (k) For the sake of convenience, the initial value of xe is set to zero. r+1 (k) The internal data v used to estimate (k) and z (k) "Internal Data v r (k) and z r (k) In this case, the internal data v (k) and z (k) The initial value of "Internal data v1 (k) and z1 (k) For example, the signal estimation unit 212 may estimate the estimated signal xe (k) The initial value of xe1 (k) is set to zero (specifically, a zero vector). The signal estimation unit 212 performs the above-mentioned initialization operation N times while updating the variable k from 1 to N. At this time, the signal estimation unit 212 may perform the N initialization operations in sequence, or may perform the N initialization operations in parallel. Then, the signal estimation unit 212 estimates the received signal y (k) , the transmission channel matrix H, and the parameter β set by the parameter setting unit 213. r(k) and the estimated signal ex r (k) Based on the received signal y (k) The correction signal ya (k) In the following, the signal estimation unit 212 generates the r-th corrected signal ya (k) "Compensation signal ya r (k) Specifically, the signal estimation unit 212 uses Equation 11 to obtain the corrected signal ya r (k) The operator "o" in Equation 11 means a Hadamard product. The signal estimation unit 212 performs the operation of step S23 described above N times while updating the variable k from 1 to N. At this time, the signal estimation unit 212 may perform the operation of step S23 N times in sequence, or may perform the operation of step S23 N times in parallel.

[0087]

number

[0088] Here, the second term on the right side of Equation 11 is the received signal y (k) A transmitted signal x that has a relatively strong correlation with (k) The transmitted signal components contained in the received signal y other than (1) From x (k-1) and x (k+1) From x (N) In other words, the second term on the right side of Equation 11 corresponds to the signal component caused by the ideal received signal y (k) The received signal y (k) Therefore, the correction signal ya (k) is essentially the sum of the received signal y and the transmitted signal x (k) It corresponds to a signal excluding information about transmission signal components other than the above (that is, interference components).

[0089] Then, the signal estimation unit 212 estimates the corrected signal ya r (k) and the submatrix H included in the transmission channel matrix H (k、k) and the parameter ρ set by the parameter setting unit 213. r (k) and internal data v r (k) and z r (k) Based on and, the estimated signal xe r+1 (k) Specifically, the signal estimation unit 212 generates the estimated signal xe using Equation 12 (step S24). r+1 (k) In addition, D(ρ r (k) ) is the parameter ρ, which is a vector containing M-dimensional vector components. r (k) As diagonal elements, the signal estimation unit 212 performs the operation of step S24 described above N times while updating the variable k from 1 to N. At this time, the signal estimation unit 212 may perform the operation of step S24 N times in sequence, or may perform the operation of step S24 N times in parallel.

[0090]

number

[0091] Then, the signal estimation unit 212 estimates the latest estimated signal xe r+1 (k) Based on the internal data v r (k) and z r (k) (step S25). That is, the signal estimation unit 212 updates the internal data v r+1 (k) and z r+1 (k) Specifically, the signal estimation unit 212 generates the internal data z r+1 (k)In addition, "Ψ(w)" in Equation 13 represents M signal components (w1, w2, . . . , w M ) is a function that performs a nonlinear transformation process on a signal (=vector) w including the r (k) The function Ψ(w) may be determined in advance according to a modulation method (particularly, a multi-value number of a signal) used in the communication system SYS. An example of the function Ψ(w) is shown in Equation 14. Equation 14 indicates that the function Ψ(w) is determined by a parameter τ r (k) (=τ r (k) #1, t r (k) #2, , τ r (k) #M) is an example of a function that specifies a nonlinear conversion process performed using a hyperbolic tangent function tanh defined by the formula (14). The function Ψ(w) shown in formula 14 is an example of a function Ψ(w) that is used when the communication system SYS uses a binary phase shift keying (BPSK (Binary Phase Shift Keying) modulation) method. The signal estimation unit 212 estimates the internal data z r+1 (k) After generating, the internal data v r+1 (k) The operator "o" in Equation 15 means a Hadamard product. The signal estimation unit 212 performs the operation of step S25 described above N times while updating the variable k from 1 to N. At this time, the signal estimation unit 212 may perform the operation of step S25 N times in sequence, or may perform the operation of step S25 N times in parallel.

[0092]

number

[0093]

number

[0094]

number

[0095] Then, the signal estimation unit 212 determines whether the variable r is smaller than a constant R used as a threshold (step S26). As described above, the variable r is determined by the signal estimation unit 212 calculating the estimated signal xe r+1 (k) Specifically, the variable r counts the number of times the signal estimation unit 212 estimates the estimated signal xe r+1 (k) The constant R indicates the number of times the signal estimation unit 212 has performed the operation of estimating the estimated signal xe r+1 (k) This is a preset value as the number of times that the operation of estimating the number of times should be performed.

[0096] As a result of the determination in step S26, when it is determined that the variable r is smaller than the constant R (step S26: Yes), the signal estimation unit 212 calculates the estimated signal xe r+1 (k) That is, the signal estimation unit 212 re-estimates the estimated signal xe r+2 (k) Specifically, the signal estimation unit 212 increments the variable r by 1 (step S28), and then performs the operations from steps S23 to S25 again to newly estimate the estimated signal xe r+1 (k) Again, the signal estimation unit 212 re-estimates the estimated signal xe using Equation 3 above. r+1 (k) The above operations from step S23 to step S25 are repeated until it is determined that the variable r is not smaller than the constant R. That is, the operations from step S23 to step S25 are repeated R times.

[0097] As a result of the determination in step S26, if it is determined that the variable r is not smaller than the constant R (step S26: No), the signal estimation unit 212 calculates the estimated signal xe r+1 (k) In this case, the signal estimation unit 212 does not need to re-estimate the last estimated signal xe r+1 (k) (=xe R (k) =xe R (0) From xe R (N-1) ) and outputs a definite estimated signal xef (step S27). Specifically, the signal estimation unit 212 uses Equation 16 to obtain the transmitted signal x (k) The final confirmed estimate signal xef (k) The symbol “Π” in Equation 16 is output (step S27). s "(w)" is a symbol for indicating a signal pattern of a transmission signal x that has the closest distance (Euclidean distance) to the signal pattern of the signal w among all signal patterns that the transmission signal x can take. In other words, the signal estimation unit 212 estimates the transmission signal x (k) Among all possible signal patterns, the estimated signal xe R (k) A transmission signal x having a signal pattern that is closest in distance (for example, Euclidean distance) to the signal pattern of (k) , the definite estimated signal xef (k) The signal estimation unit 212 performs the above-mentioned operation of step S27 N times while updating the variable k from 1 to N. At this time, the signal estimation unit 212 may perform the operation of step S27 N times in sequence, or may perform the operation of step S27 N times in parallel. As a result, the signal estimation unit 212 outputs the definite estimated signal xef (0) From the confirmed estimated signal xef (N-1) The processor 11 outputs a definite estimated signal xef (i.e., an estimate of the transmitted signal x) in which the signals are arranged in the appropriate order.

[0098]

number

[0099] <3> Technical effects of the communication system SYS As described above, the signal estimation unit 212 estimates the transmitted signal x from the received signal y using a signal estimation method whose details are defined by the parameters β, ρ, τ, and η set by the parameter setting unit 213. Here, the parameter setting unit 213 sets each of the parameters β, ρ, τ, and η to either a fixed value unrelated to the learning operation or a learned value that can be learned by the learning operation, using the classification results of the signal groups X and Y. Therefore, the signal estimation unit 212 can accurately estimate the transmitted signal x based on the received signal y.

[0100] Specifically, as described above, the parameter setting unit 213 calculates the communication capacities C1 to C3 based on the classification results of the signal groups X and Y, and sets each of the parameters β, ρ, τ, and η to either a fixed value unrelated to the learning operation or a learned value that can be learned by the learning operation based on the magnitude relationship of the communication capacities C1 to C3. Here, the magnitude relationship of the communication capacities C1 to C3 varies depending on whether the transmitting antenna 10 and the receiving antenna 20 are arranged in an ideal arrangement mode or not.

[0101] Specifically, when the communication capacity C1 and the communication capacity C2 are the same, the transmission signal x (k) The received signal y (k) Transmit signal based on x (k) The communication capacity C2 when the estimated capacity is the total capacity of all received signals y (1) From y (N) Transmit signal based on x (k) Therefore, in this case, the communication capacity C1 is the same as that when the received signal y (k) It is assumed that there is almost no signal component due to interference such as fading in the signal. In other words, it is assumed that there is a relatively high possibility that the transmitting antenna 10 and the receiving antenna 20 are arranged in an ideal arrangement. In this case, as described above, the parameter β r (k) is set to a zero vector. As a result, Equation 11 above becomes Equation 17.

[0102]

number

[0103] As shown in Equation 17, when the communication capacities C1 and C2 are the same, the signal estimation unit 212 estimates the received signal y (k) The correction signal ya (k) That is, the signal estimation unit 212 does not need to perform the operation of step S25 in FIG. 9. This is because, as described above, when the communication capacity C1 and the communication capacity C2 are the same, the received signal y (k) This is because there is a relatively high possibility that x does not include signal components caused by interference such as fading, in other words, the signal estimation unit 212 does not need to consider the second term on the right side of Equation 11, which corresponds to the signal components caused by interference such as fading. As a result, the signal estimation unit 212 can accurately estimate the transmitted signal x from the received signal y with a relatively small amount of calculation.

[0104] On the other hand, when the communication capacity C1 and the communication capacity C2 are not the same, the received signal y (k) It is assumed that there is a relatively high possibility that the signal components due to interference such as fading are included in the signal. In other words, it is assumed that there is a relatively high possibility that the transmitting antenna 10 and the receiving antenna 20 are not arranged in an ideal arrangement. In this case, as described above, the parameter β r (k) In principle, is not set to a zero vector, so the signal estimation unit 212 estimates the signal components due to interference such as fading as the received signal y (k) By performing a calculation process to eliminate (k) and the received signal y (k) Instead, the correction signal (k) Based on the estimated signal ex (k) (i.e., the transmitted signal x (k)Therefore, the signal estimation unit 212 can estimate the transmitted signal x from the received signal y, taking into account the effect of interference (i.e., taking into account that the transmitting antenna 10 and the receiving antenna 20 are not ideally arranged). As a result, the signal estimation unit 212 estimates the corrected signal ya (k) Without using the transmitted signal x (k) It is possible to estimate the transmitted signal x from the received signal y with higher accuracy compared to the case where the transmitted signal x is estimated from the received signal y.

[0105] Furthermore, the parameter β r (k) Since is set to a learnable value, the parameter β r (k) is set to a fixed value across the board, the signal estimation unit 212 estimates the signal components caused by interference such as fading in the received signal y (k) Correction signal better eliminated from ya (k) Therefore, the signal estimation unit 212 can estimate the transmitted signal x from the received signal y while more appropriately taking into account the effect of interference (i.e., after more appropriately taking into account that the transmitting antenna 10 and the receiving antenna 20 are not ideally arranged). As a result, the signal estimation unit 212 can estimate the transmitted signal x from the received signal y by using the parameter β r (k) is set to a fixed value across the board, the transmitted signal x can be estimated from the received signal y with higher accuracy.

[0106] In addition, when the communication capacity C2 and the communication capacity C3 are the same, the transmission signal x (k) The received signal y (k) Transmit signal based on x (k) The theoretical value of the communication capacity when is estimated (i.e., communication capacity C2), the transmission signal x (k) The received signal y (k) Transmit signal from x (k)is estimated. In other words, the communication capacity C3 when the transmission signal x is estimated using a signal estimation method based on the minimum mean square error coincides with its theoretical upper limit. Here, the signal estimation method based on the minimum mean square error has the advantage that the transmission signal x can be estimated with a small amount of calculation, but has the disadvantage that the estimation accuracy of the transmission signal x is significantly deteriorated under a situation in which the transmission antenna 10 and the reception antenna 20 are not arranged in an ideal arrangement. However, when the communication capacity C2 and the communication capacity C3 are the same, it is estimated that there is a relatively high possibility that the estimation accuracy of the transmission signal x is not deteriorated because the communication capacity C3 when the transmission signal x is estimated using a signal estimation method based on the minimum mean square error coincides with its theoretical upper limit. In other words, it is assumed that there is a relatively high possibility that the transmission antenna 10 and the reception antenna 20 are arranged in an ideal arrangement.

[0107] In this case, the parameter setting unit 213 sets the parameter ρ r (k) is set to a fixed value, and the parameter τ r (k) is set as the upper limit, and the parameter η r (k) is set to a zero vector. As a result, the output of the function Ψ(w) shown in the above-mentioned formula 14 is fixed to zero. This is because the parameter τ r (k) This is because, since is set to an upper limit value (e.g., a value corresponding to infinity), the argument of the hyperbolic function in Equation 14 becomes zero, and as a result, the output of the hyperbolic function becomes zero. When the output of the function Ψ(w) is fixed to zero, as shown in Equation 13, the internal data z r (k) is fixed to a zero vector. Furthermore, the parameter η r (k) When is set to a zero vector, the internal data v r (k) is also fixed to the zero vector. As a result, the estimated signal x r+1 (k) Equation 12 for estimating is expressed as Equation 18.

[0108]

number

[0109] The signal estimation method shown in Equation 18 is equivalent to a signal estimation method based on an ideal minimum mean square error (however, the correction signal ya r (k) and generating an estimated signal x r+1 (k) is performed R times). That is, when the communication capacity C2 and the communication capacity C3 are the same, the signal estimation unit 212 estimates the transmitted signal x from the received signal y using a signal estimation method based on the minimum mean square error. As a result, the signal estimation unit 212 can accurately estimate the transmitted signal x from the received signal y with a relatively small amount of calculation.

[0110] On the other hand, if the communication capacity C2 and the communication capacity C3 are not the same, it is assumed that there is a relatively high possibility that the transmitting antenna 10 and the receiving antenna 20 are not arranged in an ideal arrangement. In this case, the parameter ρ r (k) , τ r (k) and η r (k) is not set to a fixed value in principle. Therefore, the signal estimation unit 212 estimates the transmitted signal x using the signal estimation method proposed in this embodiment, which has higher estimation accuracy of the transmitted signal x than the signal estimation method based on the minimum mean square error. Therefore, the signal estimation unit 212 can estimate the transmitted signal x from the received signal y with higher accuracy than when the transmitted signal x is always estimated using the signal estimation method based on the minimum mean square error.

[0111] Furthermore, the parameter ρ r (k) , τ r (k) and η r (k) Since is set to a learnable value, the parameter ρ r(k) , τ r (k) and η r (k) Compared to a case where the parameter ρ is uniformly set to a fixed value, the signal estimation unit 212 estimates the transmission signal x using a signal estimation method tuned in consideration of the actual communication environment of the communication system SYS, etc. Therefore, the signal estimation unit 212 can estimate the transmission signal x from the reception signal y after more appropriately considering the influence of interference (i.e., after more appropriately considering that the transmission antenna 10 and the reception antenna 20 are not arranged in an ideal arrangement). As a result, the signal estimation unit 212 can estimate the transmission signal x from the reception signal y by more appropriately considering the influence of interference (i.e., after more appropriately considering that the transmission antenna 10 and the reception antenna 20 are not arranged in an ideal arrangement). r (k) , τ r (k) and η r (k) is set to a fixed value across the board, the transmitted signal x can be estimated from the received signal y with higher accuracy.

[0112] In this way, the signal estimation unit 212 can accurately estimate the transmitted signal x from the received signal y in both cases where there is a relatively high possibility that the transmitting antenna 10 and the receiving antenna 20 are arranged in an ideal arrangement, and where there is a relatively high possibility that the transmitting antenna 10 and the receiving antenna 20 are not arranged in an ideal arrangement.

[0113] In addition, when the communication capacities C1, C2, and C3 are all the same, the received signal y (k) and the correction signal ya (k) Therefore, the signal estimation method shown in Equation 18 is consistent with the signal estimation method based on the ideal minimum mean square error. Therefore, the signal estimation unit 212 estimates the corrected signal ya r (k) and generating an estimated signal x r+1 (k) In other words, the signal estimation unit 212 can accurately estimate the transmitted signal x from the received signal y with a relatively small amount of calculation. In this case, the correction signal ya r(k) and the estimated signal x r+1 (k) A constant R indicating the number of times to repeat the operation of estimating may be set to 1.

[0114] <4> Variations Next, a modification of the communication system SYS will be described.

[0115] <4-1> Modification of signal estimation operation First, a modified example of the signal estimation operation performed by the signal estimation unit 212 will be described with reference to Fig. 10. Fig. 10 is a flowchart showing the flow of the modified example of the signal estimation operation performed by the signal estimation unit 212.

[0116] As shown in FIG 10, in the modified example, the signal estimation unit 212 also performs the operations of steps S20 to S22. (1) From y (N) and the transmission channel matrix H are input, and the signal estimation unit 212 performs an initialization operation. After that, the signal estimation unit 212 sets a variable k to 1 (step S31).

[0117] Then, the signal estimation unit 212 estimates the received signal y (k) , the transmission channel matrix H, and the parameter β set by the parameter setting unit 213. r (k) and the estimated signal ex r (k) Based on the correction signal ya r (k) (Step S32). The operation of step S32 is performed by generating the above-mentioned correction signal ya r (k) Compared with the operation of step S23 of FIG. 9 for generating the correction signal ya r (k) Specifically, in the variant, the signal estimation unit 212 uses Equation 19 to generate the corrected signal ya r (k) Generate.

[0118]

number

[0119] Thereafter, in the modified example, the signal estimation unit 212 performs the operations of steps S24 to S25. That is, the signal estimation unit 212 receives the correction signal ya r (k) Using the estimated signal xe r+1 (k) Generate the internal data v r (k) and z r (k) After that, the signal estimation unit 212 determines whether the variable k is smaller than a constant N (step S33).

[0120] As a result of the determination in step S33, when it is determined that the variable k is smaller than the constant N (step S33: No), the signal estimation unit 212 calculates the estimated signal xe r+1 (1) From xe r+1 (k) On the other hand, the estimated signal xe r+1 (k+1) From xe r+1 (N) Therefore, in this case, the signal estimation unit 212 generates the estimated signal xe r+1 (k+1) From xe r+1 (N) To generate k, the variable k is incremented by 1 (step S34), and then the operations from step S23 to step S25 are repeated.

[0121] On the other hand, if it is determined in step S33 that the variable k is not smaller than the constant N (step S33: Yes), the signal estimation unit 212 calculates the estimated signal xe r+1 (1) From xe r+1 (N) In this case, it is assumed that the estimated signal xe R (1)From xe R (N) is generated (step S26: Yes), the signal estimation unit 212 r+1 (1) From xe r+1 (N) is repeated R times (steps S28, S31 to S32, S24 to S25, and S33). After that, the signal estimation unit 212 outputs the definite estimated signal xef (step S27).

[0122] As described above, in the modified example, the correction signal ya r (k) (step S32), and the operation of generating the estimated signal xe r+1 (k+1) (step S24), and the internal data v r (k) and z r (k) (step S25) is repeated N times while incrementing the variable k by 1. Here, the signal estimation unit 212 calculates the correction signal ya r (k) To this end, the signal estimation unit 212 generates the estimated signal xe r (k+1) From xe r (N) and the estimated signal xe r+1 (1) From xe r+1 (k-1) The correction signal ya generated based on r (k) Using the estimated signal xe r+1 (k) That is, the operation of step S32 in FIG. 10 generates the estimated signal xe r (k+1) From xe r (N) Not only the estimated signal xe r+1 (1) From xe r+1 (k-1) Based on the correction signal ya r (k) (As a result, the estimated signal xe r+1(k) (generating the estimated signal xe r (1) From xe r (N) Based on the correction signal ya r (k) (As a result, the estimated signal xe r+1 (k) In other words, the operation of step S32 in FIG. 10 is different from the operation of step S23 in which the estimated signal xe r+1 (1) From xe r+1 (k-1) Based on the correction signal ya r (k) (As a result, the estimated signal xe r+1 (k) (generating the estimated signal xe r+1 (1) From xe r+1 (k-1) Without using the correction signal ya r (k) (As a result, the estimated signal xe r+1 (k) As a result, in the modified example, the estimated signal ex is generated in a step S23 in comparison with the above-described operation shown in FIG. r+1 (k) is the transmitted signal x (k) In the modified example, the time required for the estimated signal xe to converge to is relatively high. r+1 (k) The estimated signal xe generated before incrementing the variable r to generate r (1) From xe r (N) , as well as the estimated signal xe generated after incrementing the variable r. r+1 (1) From xe r+1 (N) That is, in the modified example, the signal estimation unit 212 obtains the above-mentioned various effects while estimating the estimated signal ex r+1 (k) is the transmitted signal x (k) This can shorten the time it takes for the signal to converge to a given value.

[0123] <4-2> Modifications of the transmitting antenna 10 and the receiving antenna 20 Next, modified examples of the transmitting antenna 10 and the receiving antenna 20 will be described. In the above description, the communication system SYS includes the transmitting antenna 10 in which M array antennas 100 are arranged concentrically. On the other hand, in a modified example, the communication system SYS may include the transmitting antenna 10a in which M array antennas 100 are arranged linearly as shown in FIG. 11. Similarly, in the above description, the communication system SYS includes the receiving antenna 20 in which M array antennas 200 are arranged concentrically. On the other hand, in a modified example, the communication system SYS may include the receiving antenna 20a in which M array antennas 200 are arranged linearly as shown in FIG. 11. The transmitting antenna 10a and the receiving antenna 20a of such a modified example are superior in installation properties compared to the transmitting antenna 10 and the receiving antenna 20.

[0124] <5> Additional Notes The following supplementary notes are further disclosed regarding the above-described embodiment. [Appendix 1] A transmitting device that transmits a transmission signal using a transmitting antenna including a plurality of transmitting antenna elements; a receiving device that receives the transmission signal transmitted by the transmitting device as a received signal using a receiving antenna including a plurality of receiving antenna elements; In a communication system comprising: a signal estimation device for estimating the transmission signal from the reception signal, The signal estimation device comprises: a parameter setting means for classifying a plurality of reception signal components contained in the reception signal and received by the plurality of reception antenna elements respectively and a plurality of transmission signal components contained in the transmission signal and transmitted by the plurality of transmission antenna elements respectively into a plurality of signal groups according to the strength of correlation between the reception signal components and the transmission signal components, based on system information including antenna information on the transmission antenna and the reception antenna, distance information on a transmission distance between the transmission antenna and the reception antenna, carrier information on a carrier used by the communication system, and processing information on signal processing performed by the transmission device and reception processing performed by the reception device, and for setting a parameter defining the content of a signal estimation method for each of the signal groups to either a learning value that can be learned by a learning operation or a fixed value that is unrelated to the learning operation; a signal estimation means for estimating, for each of the signal groups, an estimated signal that is an estimated value of the transmitted signal from the received signal by using the signal estimation method defined by the parameters set by the parameter setting means; A signal estimation device comprising: [Appendix 2] The parameter setting means estimating a transmission path matrix indicating a state of a transmission path between the transmitting device and the receiving device based on the system information; classifying the plurality of received signal components and the plurality of transmitted signal components into the plurality of signal groups by using the magnitude of the absolute value of each component of the transmission channel matrix as a correlation strength between one transmitted signal component and one received signal component corresponding to each component; Based on the transmission channel matrix and the classification results of the plurality of signal groups, (i) a first communication capacity indicating a transmittable amount of information when each transmission signal component is estimated from the plurality of received signal components, (ii) a second communication capacity indicating a theoretical value of a transmittable amount of information when each transmission signal component is estimated from a received signal component having a relatively strong correlation with each transmission signal component, and (iii) a third communication capacity indicating a transmittable amount of information when each transmission signal component is estimated from a received signal component having a relatively strong correlation with each transmission signal component using a signal estimation method other than the signal estimation method used by the signal estimation means, The parameter is set to either the learned value or the fixed value based on a magnitude relationship between the first to third communication capacities. 2. A signal estimation device as claimed in claim 1. [Appendix 3] the parameter setting means sets each of the plurality of parameters to either the learned value or the fixed value; the parameter setting means sets a first parameter of the plurality of parameters to the fixed value when the first communication capacity and the second communication capacity are equal, and sets the first parameter to the learned value when the first communication capacity and the second communication capacity are not equal; The parameter setting means sets a second parameter, which is different from the first parameter, of the plurality of parameters to the fixed value when the second communication capacity and the third communication capacity are equal, and sets the second parameter to the learned value when the first communication capacity and the second communication capacity are not equal. 3. A signal estimation device according to claim 2. [Appendix 4] The received signal component y , which is classified into the kth (k is a variable indicating an integer between 1 and N, and N is a constant of 2 or more indicating the number of the signal groups)-th signal group among the plurality of signal groups. (k) The transmission signal component x (k) The estimated signal xe (k) To estimate The estimated signal xe (k) The internal data v used to estimate (k) and z (k) and the estimated signal xe (k) A first operation is performed to initialize After performing the first operation, the received signal component y (k) and a part of a transmission path matrix H indicating the state of a transmission path between the transmitting device and the receiving device and the received signal component y (k) while the transmitted signal component x (k)The submatrix H that does not contain the element corresponding to (k、p) (p is a variable indicating an integer between 1 and N, and different from the variable k), and the estimated signal xe (0) From xe (k-1) and xe (k+1) From xe (N-1) Based on the received signal component y (k) The correction signal ya (k) and performing a second operation to generate After performing the second operation, the correction signal ya is estimated by using the signal estimation method determined based on the parameters set by the parameter setting means. (k) and a part of the transmission channel matrix H and the received signal component y (k) and the transmitted signal component x (k) A submatrix H containing elements corresponding to (k、k) and the internal data v (k) and z (k) and based on the estimated signal xe (k) By updating the estimated signal xe (k) The third step is to re-estimate After repeating the second and third operations R times (R is a constant indicating an integer equal to or greater than 2), the transmission signal component x (k) Among multiple possible signal component candidates, the estimated signal xe (k) outputting a signal component candidate that is closest to When the second operation is performed r times, the signal estimation means estimates the estimated signal xe (0) From xe (k-1) and xe (k+1) From xe (N-1) Based on the correction signal ya (k) Generate 4. A signal estimation device according to any one of claims 1 to 3. [Appendix 5] The received signal component y , which is classified into the kth (k is a variable indicating an integer between 1 and N, and N is a constant of 2 or more indicating the number of the signal groups)-th signal group among the plurality of signal groups. (k) The transmission signal component x(k) The estimated signal xe (k) To estimate The estimated signal xe (1) From xe (N) The internal data v used to estimate (1) From v (N) and z (1) From z (N) and the estimated signal xe (1) From xe (N) A first operation is performed to initialize After performing the first operation, the received signal component y (k) and a part of a transmission path matrix H indicating the state of a transmission path between the transmitting device and the receiving device and the received signal component y (k) while the transmitted signal component x (k) The submatrix H that does not contain the element corresponding to (k、p) (p is a variable indicating an integer between 1 and N, and different from the variable k), and the estimated signal xe (0) From xe (k-1) and xe (k+1) From xe (N-1) Based on the received signal component y (k) The correction signal ya (k) and generating the correction signal ya using the signal estimation method determined based on the parameters set by the parameter setting means. (k) and a part of the transmission channel matrix H and the received signal component y (k) and the transmitted signal component x (k) A submatrix H containing elements corresponding to (k、k) and the internal data v (k) and z (k) and based on the estimated signal xe (k) By updating the estimated signal xe (k) A second operation is performed in which a predetermined process of re-estimating is repeated N times while updating the variable k from 1 to N; After repeating the second operation R times (R is a constant indicating an integer equal to or greater than 2), the transmission signal component x (k)Among multiple possible signal component candidates, the estimated signal xe (k) outputting a signal component candidate that is closest to When the k-th predetermined process is performed in the r-th (r is a variable indicating an integer not less than 1 and not more than R) second operation, the signal estimation means estimates the estimated signal xe estimated by the k+1-th to N-th predetermined processes in the r-1-th second operation. (k+1) From xe (N) and the estimated signal xe estimated by the 1st to k-1th predetermined processing in the rth second operation. (1) From xe (k-1) Based on the correction signal ya (k) Calculate 4. A signal estimation device according to any one of claims 1 to 3. [Appendix 6] The transmitting antenna is an antenna in which M (where M is an integer equal to or greater than 1) transmitting array antennas are arranged concentrically, each antenna including N (where N is a constant representing an integer equal to or greater than 2) transmitting antenna elements arranged at equal intervals on a circle, The receiving antenna is an antenna in which M receiving array antennas, each of which includes N receiving antenna elements arranged at equal intervals on a circle, are arranged concentrically. 6. A signal estimation device according to any one of claims 1 to 5. [Appendix 7] The transmitting antenna is an antenna in which M (where M is an integer equal to or greater than 1) transmitting array antennas are arranged in a line, each antenna including N (where N is a constant representing an integer equal to or greater than 2) transmitting antenna elements arranged at equal intervals on a circle, The receiving antenna is an antenna in which M receiving array antennas, each of which includes N receiving antenna elements arranged at equal intervals on a circle, are arranged in a straight line. 6. A signal estimation device according to any one of claims 1 to 5. [Appendix 8] The receiving device further includes a transform unit for performing a discrete Fourier transform process on the plurality of received signal components in units of a predetermined number of the received signal components as the signal processing, The signal estimation means estimates the estimated signal from the received signal that has been subjected to the discrete Fourier transform processing. 8. A signal estimation device according to any one of claims 1 to 7. [Appendix 9] The learning operation includes an operation of updating the parameters based on learning data including sample values ​​of the transmission signal and sample values ​​of the reception signal so that a loss function relating to an error between the estimated signal estimated based on the sample values ​​of the reception signal and the signal estimation method and the sample values ​​of the transmission signal is minimized. 9. A signal estimation device according to any one of claims 1 to 8. [Appendix 10] A transmitting device that transmits a transmission signal using a transmitting antenna including a plurality of transmitting antenna elements; a receiving device that receives the transmission signal transmitted by the transmitting device as a received signal using a receiving antenna including a plurality of receiving antenna elements; A signal estimation method for estimating a transmission signal from a received signal in a communication system comprising: The signal estimation method includes: based on system information including antenna information on the transmitting antenna and the receiving antenna, distance information on a transmission distance between the transmitting antenna and the receiving antenna, carrier information on a carrier used by the communication system, and processing information on signal processing performed by the transmitting device and reception processing performed by the receiving device, classifying a plurality of received signal components included in the received signal and received by the plurality of receiving antenna elements, and a plurality of transmitted signal components included in the transmitted signal and transmitted by the plurality of transmitting antenna elements, into a plurality of signal groups according to the strength of correlation between the received signal components and the transmitted signal components; setting a parameter defining the content of a signal estimation method for each of the signal groups to either a learning value that can be learned by a learning operation or a fixed value that is unrelated to the learning operation; Estimating an estimated signal, which is an estimate of the transmitted signal, from the received signal for each of the signal groups using the signal estimation method defined by the set parameters. Signal estimation method. [Appendix 11] A computer program for causing a computer to execute a signal estimation method, comprising: The signal estimation method includes: A transmitting device that transmits a transmission signal using a transmitting antenna including a plurality of transmitting antenna elements; a receiving device that receives the transmission signal transmitted by the transmitting device as a received signal using a receiving antenna including a plurality of receiving antenna elements; A signal estimation method for estimating a transmission signal from a received signal in a communication system comprising: The signal estimation method includes: based on system information including antenna information on the transmitting antenna and the receiving antenna, distance information on a transmission distance between the transmitting antenna and the receiving antenna, carrier information on a carrier used by the communication system, and processing information on signal processing performed by the transmitting device and reception processing performed by the receiving device, classifying a plurality of received signal components included in the received signal and received by the plurality of receiving antenna elements, and a plurality of transmitted signal components included in the transmitted signal and transmitted by the plurality of transmitting antenna elements, into a plurality of signal groups according to the strength of correlation between the received signal components and the transmitted signal components; setting a parameter defining the content of a signal estimation method for each of the signal groups to either a learning value that can be learned by a learning operation or a fixed value that is unrelated to the learning operation; Estimating an estimated signal, which is an estimate of the transmitted signal, from the received signal for each of the signal groups using the signal estimation method defined by the set parameters. Computer program.

[0125] The present invention can be modified as appropriate within the scope that does not deviate from the gist or concept of the invention that can be read from the claims and the entire specification, and the signal estimation device, signal estimation method, and computer program involving such modifications are also included in the technical concept of the present invention. [Explanation of symbols]

[0126] 1. Transmitting device 11 Signal Processing Device 2. Receiving device 21 Signal Processing Device 211 Conversion Unit 212 Signal Estimation Unit 213 Parameter Setting Unit SYS Communication System x Transmission signal y Received signal

Claims

1. A transmitting device that transmits a transmission signal using a transmitting antenna including a plurality of transmitting antenna elements; a receiving device that receives the transmission signal transmitted by the transmitting device as a received signal using a receiving antenna including a plurality of receiving antenna elements; In a communication system comprising: a signal estimation device for estimating the transmission signal from the reception signal, The signal estimation device comprises: a parameter setting means for classifying a plurality of reception signal components contained in the reception signal and received by the plurality of reception antenna elements respectively and a plurality of transmission signal components contained in the transmission signal and transmitted by the plurality of transmission antenna elements respectively into a plurality of signal groups according to the strength of correlation between the reception signal components and the transmission signal components, based on system information including antenna information on the transmission antenna and the reception antenna, distance information on a transmission distance between the transmission antenna and the reception antenna, carrier information on a carrier used by the communication system, and processing information on signal processing performed by the transmission device and reception processing performed by the reception device, and for setting a parameter defining the content of a signal estimation method for each of the signal groups to either a learning value that can be learned by a learning operation or a fixed value that is unrelated to the learning operation; a signal estimation means for estimating, for each of the signal groups, an estimated signal that is an estimated value of the transmitted signal from the received signal by using the signal estimation method defined by the parameters set by the parameter setting means; A signal estimation device comprising:

2. The parameter setting means estimating a transmission path matrix indicating a state of a transmission path between the transmitting device and the receiving device based on the system information; classifying the plurality of received signal components and the plurality of transmitted signal components into the plurality of signal groups by using the magnitude of the absolute value of each component of the transmission channel matrix as a correlation strength between one transmitted signal component and one received signal component corresponding to each component; Based on the transmission path matrix and the classification results of the plurality of signal groups, (i) a first communication capacity indicating an amount of information that can be transmitted when each transmission signal component is estimated from the plurality of received signal components, (ii) a second communication capacity indicating a theoretical value of an amount of information that can be transmitted when each transmission signal component is estimated from a received signal component having a relatively strong correlation with each transmission signal component, and (iii) a third communication capacity indicating an amount of information that can be transmitted when each transmission signal component is estimated from a received signal component having a relatively strong correlation with each transmission signal component using a signal estimation method other than the signal estimation method used by the signal estimation means, The parameter is set to either the learned value or the fixed value based on a magnitude relationship between the first to third communication capacities.

2. A signal estimation device according to claim 1.

3. the parameter setting means sets each of the plurality of parameters to either the learned value or the fixed value; the parameter setting means sets a first parameter of the plurality of parameters to the fixed value when the first communication capacity and the second communication capacity are equal, and sets the first parameter to the learned value when the first communication capacity and the second communication capacity are not equal; The parameter setting means sets a second parameter, which is different from the first parameter, of the plurality of parameters to the fixed value when the second communication capacity and the third communication capacity are equal, and sets the second parameter to the learned value when the first communication capacity and the second communication capacity are not equal.

3. A signal estimation device according to claim 2.

4. The received signal component y is classified into the k-th signal group (k is a variable indicating an integer of 1 or more and N or less, and N is a constant of 2 or more indicating the number of the signal groups) among the plurality of signal groups. （ｋ） from the transmission signal component x classified into the k-th signal group （ｋ） The estimated signal xe （ｋ） To estimate The estimated signal xe （ｋ） The internal data vr used to estimate （ｋ） and zr （ｋ） and the estimated signal xe （ｋ） performing a first operation for initializing After performing the first operation, the received signal component y is estimated by using the signal estimation method determined based on the parameters set by the parameter setting means. （ｋ） and a part of a transmission path matrix H indicating the state of a transmission path between the transmitting device and the receiving device and the received signal component y （ｋ） while the transmitted signal component x （ｋ） A submatrix H that does not contain an element corresponding to （ｋ、ｐ） (p is a variable indicating an integer greater than or equal to 1 and less than or equal to N, and different from the variable k), and the estimated signals xe(1) to xe （ｋ－１） and xe （ｋ＋１） Based on xe(N), the received signal component y （ｋ） The correction signal ya （ｋ） and performing a second operation of generating After performing the second operation, the correction signal ya is calculated using the signal estimation method determined based on the parameters set by the parameter setting means. （ｋ） and a part of the transmission channel matrix H and the received signal component y （ｋ） and the transmitted signal component x （ｋ） A submatrix H containing elements corresponding to （ｋ、ｋ） and the internal data vr （ｋ） and zr （ｋ） Based on the above, the estimated signal xe （ｋ） By updating the estimated signal xe （ｋ） A third operation is performed to re-estimate After repeating the second and third operations R times (R is a constant indicating an integer equal to or greater than 2), the transmission signal component x （ｋ） Among a plurality of possible signal component candidates, the estimated signal xe （ｋ） outputting a signal component candidate that is closest to When the second operation is performed r+1 times (r is a variable indicating an integer equal to or greater than 1 and less than R), the signal estimation means subtracts the estimated signal xe(1) from the estimated signal xe(2) estimated by the third operation performed r times. （ｋ－１） and xe （ｋ＋１） Based on xe(N), the correction signal ya （ｋ） Generate Internal Data ｒ （ｋ） and z ｒ （ｋ） is a value for estimating the estimated signal xe(k), which is stored and used to estimate the estimated signal xe(k), and the internal data z ｒ＋１ （ｋ） is the estimated signal xe(k) and the internal data v ｒ （ｋ） and the internal data v ｒ＋１ （ｋ） is the estimated signal xe(k) and the internal data z ｒ＋１ （ｋ） is generated using The submatrix H （ｋ、ｐ） and the submatrix H （ｋ、ｋ） The first variable is the received signal component y （ｋ） is the index of the classified signal group, and the second variable is the transmitted signal component x （ｋ） is the index of the classified signal group, and the submatrix H （ｋ、ｐ） is the received signal component y （ｋ） The index of the signal group into which the transmitted signal component x （ｋ） represents a case where the index of the classified signal group is different from that of the submatrix H （ｋ、ｋ） is the received signal component y （ｋ） The index of the signal group into which the transmitted signal component x （ｋ） represents the case where the index of the classified signal group is the same as A signal estimation device according to any one of claims 1 to 3.

5. The received signal component y is classified into the k-th signal group (k is a variable indicating an integer of 1 or more and N or less, and N is a constant of 2 or more indicating the number of the signal groups) among the plurality of signal groups. （ｋ） from the transmission signal component x classified into the k-th signal group （ｋ） The estimated signal xe （ｋ） To estimate The estimated signal xe （１） From xe （Ｎ） The internal data vr used to estimate （１） From vr （Ｎ） and zr （１） From zr （Ｎ） and the estimated signal xe （１） From xe （Ｎ） performing a first operation to initialize After performing the first operation, the received signal component y is estimated by using the signal estimation method determined based on the parameters set by the parameter setting means. （ｋ） and a part of a transmission path matrix H indicating the state of a transmission path between the transmitting device and the receiving device and the received signal component y （ｋ） while the transmitted signal component x （ｋ） A submatrix H that does not contain an element corresponding to （ｋ、ｐ） (p is a variable indicating an integer greater than or equal to 1 and less than or equal to N, and different from the variable k), and the estimated signals xe(1) to xe （ｋ－１） and xe （ｋ＋１） Based on xe(N), the received signal component y （ｋ） The correction signal ya （ｋ） and generating the correction signal ya using the signal estimation method determined based on the parameters set by the parameter setting means. （ｋ） and a part of the transmission channel matrix H and the received signal component y （ｋ） and the transmitted signal component x （ｋ） A submatrix H containing elements corresponding to （ｋ、ｋ） and the internal data vr （ｋ） and zr （ｋ） Based on the above, the estimated signal xe （ｋ） By updating the estimated signal xe （ｋ） A second operation is performed in which a predetermined process of re-estimating is repeated N times while updating the variable k from 1 to N; After repeating the second operation R times (R is a constant indicating an integer equal to or greater than 2), the transmission signal component x （ｋ） Among a plurality of possible signal component candidates, the estimated signal xe （ｋ） outputting a signal component candidate that is closest to When the predetermined process is performed for the kth time in the r+1th (r is a variable indicating an integer equal to or greater than 1 and less than R) second operation, the signal estimation means estimates the estimated signal xe estimated by the k+1th to Nth predetermined processes in the rth second operation. （ｋ＋１） From xe （Ｎ） and the estimated signal xe estimated by the 1st to k−1th predetermined processing in the r+1th second operation. （１） From xe （ｋ－１） Based on the correction signal ya （ｋ） Calculate Internal Data ｒ （ｋ） and z ｒ （ｋ） is a value for estimating the estimated signal xe(k), which is stored and used to estimate the estimated signal xe(k), and the internal data z ｒ＋１ （ｋ） is the estimated signal xe(k) and the internal data v ｒ （ｋ） and the internal data v ｒ＋１ （ｋ） is the estimated signal xe(k) and the internal data z ｒ＋１ （ｋ） is generated using The submatrix H （ｋ、ｐ） and the submatrix H （ｋ、ｋ） The first variable is the received signal component y （ｋ） is the index of the classified signal group, and the second variable is the transmitted signal component x （ｋ） is the index of the classified signal group, and the submatrix H （ｋ、ｐ） is the received signal component y （ｋ） The index of the signal group into which the transmitted signal component x （ｋ） represents a case where the index of the classified signal group is different from that of the submatrix H （ｋ、ｋ） is the received signal component y （ｋ） The index of the signal group into which the transmitted signal component x （ｋ） represents the case where the index of the classified signal group is the same as A signal estimation device according to any one of claims 1 to 3.

6. the transmitting antenna is an antenna in which M (where M is an integer equal to or greater than 1) transmitting array antennas are arranged concentrically, each antenna including N (where N is a constant representing an integer equal to or greater than 2) transmitting antenna elements arranged at equal intervals on a circle, The receiving antenna is an antenna in which M receiving array antennas, each of which includes N receiving antenna elements arranged at equal intervals on a circle, are arranged concentrically. A signal estimation device according to any one of claims 1 to 3.

7. the transmitting antenna is an antenna in which M (where M is an integer equal to or greater than 1) transmitting array antennas are arranged concentrically, each antenna including the N transmitting antenna elements arranged at equal intervals on a circle, The receiving antenna is an antenna in which M receiving array antennas, each of which includes N receiving antenna elements arranged at equal intervals on a circle, are arranged concentrically.

6. A signal estimation device according to claim 4 or 5.

8. the transmitting antenna is an antenna in which M (wherein M is an integer equal to or greater than 1) transmitting array antennas are linearly arranged, the M (wherein M is an integer equal to or greater than 1) transmitting array antennas including N (wherein N is a constant equal to or greater than 2) transmitting antenna elements arranged at equal intervals on a circle, The receiving antenna is an antenna in which M receiving array antennas, each of which includes N receiving antenna elements equally spaced in a circle, are arranged in a straight line. A signal estimation device according to any one of claims 1 to 3.

9. the transmitting antenna is an antenna in which M (where M is an integer equal to or greater than 1) transmitting array antennas are linearly arranged, the M including the N transmitting antenna elements equally spaced apart in a circle, The receiving antenna is an antenna in which M receiving array antennas, each of which includes N receiving antenna elements equally spaced in a circle, are arranged in a straight line.

6. A signal estimation device according to claim 4 or 5.

10. The receiving device further includes a transform unit for performing a discrete Fourier transform process on the plurality of received signal components in units of a predetermined number of the received signal components as the receiving process, The signal estimation means estimates the estimated signal from the received signal that has been subjected to the discrete Fourier transform processing. A signal estimation device according to any one of the preceding claims.

11. The learning operation includes an operation of updating the parameters based on learning data including sample values ​​of the transmission signal and sample values ​​of the reception signal so that a loss function relating to an error between the estimated signal estimated based on the sample values ​​of the reception signal and the signal estimation method and the sample values ​​of the transmission signal is minimized. A signal estimation device according to any one of the preceding claims.

12. A transmitting device that transmits a transmission signal using a transmitting antenna including a plurality of transmitting antenna elements; a receiving device that receives the transmission signal transmitted by the transmitting device as a received signal using a receiving antenna including a plurality of receiving antenna elements; A signal estimation method for estimating a transmission signal from a received signal in a communication system comprising: The signal estimation method includes: based on system information including antenna information on the transmitting antenna and the receiving antenna, distance information on a transmission distance between the transmitting antenna and the receiving antenna, carrier information on a carrier used by the communication system, and processing information on signal processing performed by the transmitting device and reception processing performed by the receiving device, classifying a plurality of received signal components included in the received signal and received by the plurality of receiving antenna elements, and a plurality of transmitted signal components included in the transmitted signal and transmitted by the plurality of transmitting antenna elements, into a plurality of signal groups according to the strength of correlation between the received signal components and the transmitted signal components; setting a parameter defining the content of a signal estimation method for each of the signal groups to either a learning value that can be learned by a learning operation or a fixed value that is unrelated to the learning operation; Estimating an estimated signal, which is an estimate of the transmitted signal, from the received signal for each of the signal groups using the signal estimation method defined by the set parameters. Signal estimation method.

13. A computer program for causing a computer to execute a signal estimation method, comprising: The signal estimation method includes: A transmitting device that transmits a transmission signal using a transmitting antenna including a plurality of transmitting antenna elements; a receiving device that receives the transmission signal transmitted by the transmitting device as a received signal using a receiving antenna including a plurality of receiving antenna elements; A signal estimation method for estimating a transmission signal from a received signal in a communication system comprising: The signal estimation method includes: based on system information including antenna information on the transmitting antenna and the receiving antenna, distance information on a transmission distance between the transmitting antenna and the receiving antenna, carrier information on a carrier used by the communication system, and processing information on signal processing performed by the transmitting device and reception processing performed by the receiving device, classifying a plurality of received signal components included in the received signal and received by the plurality of receiving antenna elements, and a plurality of transmitted signal components included in the transmitted signal and transmitted by the plurality of transmitting antenna elements, into a plurality of signal groups according to the strength of correlation between the received signal components and the transmitted signal components; setting a parameter defining the content of a signal estimation method for each of the signal groups to either a learning value that can be learned by a learning operation or a fixed value that is unrelated to the learning operation; Using the signal estimation method defined by the set parameters, an estimated signal that is an estimate of the transmitted signal is estimated from the received signal for each of the signal groups. Computer program.

Images