JP4584155B2 - Wireless communication method and wireless communication apparatus - Google Patents

Description

  The present invention uses the same frequency channel, spatially multiplexes and transmits independent signal sequences from a plurality of different transmission antennas, receives signals using a plurality of reception antennas, and obtains a channel response matrix between the transmission and reception antennas. The number of spatially multiplexed signal sequences in a high-speed wireless access system that uses MIMO (Multiple-Input Multiple-Output) communication that demodulates data on the receiving station side and simultaneously transmits information to multiple communication partners The present invention relates to a wireless communication method and a wireless communication apparatus provided with the above antenna.

  In recent years, the IEEE802.11g standard, the IEEE802.11a standard, and the like are remarkable as high-speed wireless access systems using the 2.4 GHz band or the 5 GHz band. In these systems, an orthogonal frequency division multiplexing (OFDM) modulation method, which is a technique for stabilizing characteristics in a multipath fading environment, is used, and a transmission rate of 54 Mbps at the maximum is realized. However, the transmission rate here is a transmission rate on the physical layer. Actually, since the transmission efficiency in the MAC (Medium Access Control) layer is about 50 to 70%, the upper limit value of the actual throughput is It is about 30 Mbps. On the other hand, in the world of wired LANs, the provision of 100 Mbps high-speed lines has become widespread due to the widespread use of Ethernet's 100Base-T interface and FTTH (Fiber to the home) using optical fibers in homes. In the world of LAN, further increase in transmission speed is demanded.

  As a technology for that purpose, the MIMO technology is promising. This MIMO technology is such that different independent signals are transmitted on the same channel from a plurality of transmitting antennas on the transmitting station side, and signals are received using the same plurality of antennas on the receiving station side, between each transmitting antenna / receiving antenna. The channel response matrix is obtained, the independent signal transmitted from each antenna is estimated on the transmitting station side using this matrix, and the data is reproduced.

The directivity control that is optimal for a single user is shown, where M _{T is} the number of antenna elements of the transmitting device, M _{R is} the number of antenna elements of the receiving device that is the communication partner, and L is the number of communication sequences transmitted simultaneously in the same frequency band. . FIG. 18 is a configuration example of a transmission unit in the prior art that controls transmission directivity so as to be optimal for the propagation environment and improves the transmission rate by spatial multiplexing. Code 900 data dividing circuit, 901-1~901-L modulation circuit, the transmission signal conversion circuit 902, _{903-1~903-M T} wireless unit, _{904-1~904-M T} is the antenna element, 905 Is a transmission weight calculation circuit, and 906 is a channel response matrix acquisition circuit.

The antennas 904-1 to 904 -M _T and the radio units 903-1 to 903 -M _T can transmit and receive radio signals, and via these antennas 904-1 to 904 -M of the transmission unit. _A channel response matrix acquisition circuit 906 can estimate a channel response matrix between _T and each antenna of the communication partner. Although this method obtains the channel response matrix is not specified here, the feedback included in the estimate based on whether or received signal information obtained when performing the reception of the known signal in the antenna 904-1~904-M _T Information of the channel response matrix is acquired by information included in the information. This information is input to the transmission weight calculation circuit 905, and the transmission weight at each antenna of each signal series is calculated.

Next, when data is input to be transmitted, the data dividing circuit 900 divides one signal into L signal sequences and inputs the signals to the modulation circuits 901-1 to 901-L. Here, a preamble signal or the like for MIMO channel estimation is given and modulated, and these signals are input to the transmission signal conversion circuit 902. Here, the transmission data is multiplied by a transmission weight, is input to the radio section _{903-1~903-M T,} is transmitted as a radio signal via the antenna _{904-1~904-M T.}

The channel response matrix H (M _R × M _T matrix) obtained in the channel response matrix acquisition circuit 906 is subjected to a singular value decomposition as shown in the following equation, so that a unitary matrix V (M _T × M _T matrix), U _U (M _R × and M _R matrix) and eigenvalues √λ diagonal elements, the non-diagonal matrix can be divided into 0 to the matrix D (M _R × M _T matrix).

Here, H _ij represents a transfer coefficient from the j-th antenna of the transmission apparatus to the i-th antenna of the reception apparatus, and V _ij is a transmission weight applied to the i-th antenna element for the j-th transmission beam in the transmission apparatus. U _ij is a complex conjugate of the reception weight applied to the reception signal of the i-th antenna for the j-th transmission beam of the receiving apparatus. Here, the eigenvalue λ represents the size of the transmission capacity of each path (λ ₁ ≧ λ ₂ ≧... ≧ λ _MR ). The superscript H represents a conjugate complex matrix.

From the matrix V obtained in this way, the uplink transmission weight W obtained by selecting the column vector by the spatial multiplexing number L used for communication from the corresponding large eigenvalue is set as the transmission weight of the transmission device, and U ^{H is used} for communication. By using the uplink reception weight W ′ obtained by selecting L row vectors to be used as the reception weight of the reception device, the maximum transmission capacity corresponding to the singular value λ can be realized in each signal. W and W ′ are shown in the following equation.

When L = M _R , the transmission signal is transmitted to the transmission signal S (M _R × 1 vector) by using the transmission weight V, and the reception signal X (M _R × 1 vector) can be expressed as follows. .

  Therefore, the transmission signal S can be obtained by multiplying the reception signal X by, for example, a conjugate complex transpose matrix of U, to obtain the transmission signal S multiplied by the square root of the corresponding eigenvalue. The ratio (SN ratio) with respect to the signal becomes high, and communication with the maximum transmission capacity can be realized.

The above is a case where there is one communication partner. Next, consider a case where there are a plurality of communication partners. In the following, the number of communication partners simultaneously communicating is M _U , the number of transmission antenna elements is M _T , the number of reception antenna elements of the kth communication partner is M _R (k), and the total reception antennas of the M _U communication partners _Let M _{R be} the number, L (k) be the number of transmission communication sequences for the k-th communication partner, and L (L ≦ M _T ) be the total number of communication sequences to be transmitted.

The channel response matrix expressed by Equation (1) is defined for each communication partner, and the channel response matrix H _(k) (M _R (k) × M _T matrix) for the kth communication partner can be expressed as follows.

Further, the total spatial channel response matrix H _all (M _R × M _T matrix) is expressed as [H ₍₁₎ ^T _, H ₍₂₎ ^T _, ..., H _(Mu) ^T _, ] ^T
It is defined as

Also, the transmission weight used for each communication partner is W (k) (M _T × L (k) matrix), the transmission signal is X (k) (L (k) × 1 vector), and the reception signal is Y (k). ) (L (k) × 1 vector), the whole transmission weight is W _all
W _all = [W _(1), W _(2), ..., W _(Mu), ] (M _T × L matrix),
X _all
X _all = [X ₍₁₎ ^T _, X ₍₂₎ ^T _, ..., X _(Mu) ^T _, ] ^T (L × 1 vector), the entire received signal is Y _{all
^{Y all = [Y (1)}} T, Y (2) T, ···, Y (Mu) T,] and T _{(M R} × 1 vector). Then, the received signal Y _all can be expressed as follows.

In order to clarify a problem in communication with a plurality of communication partners, L = M _T , W _all is a diagonal matrix with diagonal elements as 1 and non-diagonal elements as 0, and a transmission apparatus Considering the case of transmitting non-directionally, it is as follows.

Here, Z _all represents a thermal noise vector, and Z _(k) is a thermal noise vector corresponding to the received signal of each communication partner. If attention is paid only to the communication partner 1, the received signal Y ₍₁₎ can be expressed as follows.

Therefore, in addition to the transmission signal X ₍₁₎ transmitted to the communication partner 1, X _{(2) to} X _(Mu) are received as interference signals, greatly degrading the transmission quality, or substantially decoding. Will not be able to.
(Miyashita, K .; Nishimura, T .; Ohgane, T .; Ogawa, Y .; Takatori, Y .; Keizo Cho) channel, '' Vehicular Technology Conference, 2002. Proceedings. VTC 2002-Fall. 2002 IEEE 56th, Volume: 3, 24-28 Sep. 2002 Pages: 13 02_1306 vol.

  In the above-described conventional example, it is possible to obtain the maximum transmission capacity for a single communication partner, but when there are a plurality of communication partners at the same time and transmission is performed to these communication partners at the same time, very large interference is caused. It will occur. This is a great detrimental effect in increasing the effect of space division multiplexing by aiming at a higher transmission rate by increasing the number of communication partners that communicate at the same time.

  The present invention has been made in view of such circumstances, and when there are a plurality of communication partners, the transmission weight is set so as to realize a desired quality by controlling an interference signal to other than the desired communication partner. It is an object to provide a wireless communication method and a wireless communication apparatus for determination.

The present invention according to claim 1 includes a plurality of transmission antenna elements, and is provided between a reception antenna and a transmission antenna element included in a communication partner, or between a reception beam formed using the reception antenna and a transmission antenna. A wireless communication method for estimating a channel response matrix representing a propagation environment of the transmission, determining a transmission weight suitable for the propagation environment, performing transmission weighting on the transmission signal, and transmitting the transmission signal to a plurality of communication partners, For each communication partner that performs communication, from the channel response vector group between the reception antenna or reception beam of the communication partner other than the communication partner and the transmission antenna of the transmission device, determining the interference space basis vector group of the communication partner; Computes an orthogonal spatial channel response matrix that extracts the component orthogonal to the interference spatial basis vector from the channel response matrix A step of determining a transmission weight obtained by a linear operation from the orthogonal spatial channel response matrix, and an orthogonalization method to a row vector group of a complex conjugate matrix of a channel response matrix of a communication partner as an interference space basis vector candidate Calculating a unit vector obtained by using a singular value decomposition of a channel response matrix of a communication partner or a vector obtained by QR decomposition, and a step related to a different communication partner Inner product of vectors
P _{kl, ab} = (ω _{k, a} , ω _{l, b} )
(K ≠ _l, ω _{k, a} is the a th vector for the k th communication partner, and ω _{l, b} is the b th vector for the l th communication partner.)
And a vector is selected so that the inner product value becomes low, and an interference space basis vector is determined.

The present invention according to claim 2 includes a plurality of transmitting antenna elements and is provided between a receiving antenna and a transmitting antenna element included in a communication partner, or between a receiving beam and a transmitting antenna formed by using the receiving antenna. A wireless communication method for estimating a channel response matrix representing a propagation environment of the transmission, determining a transmission weight suitable for the propagation environment, performing transmission weighting on the transmission signal, and transmitting the transmission signal to a plurality of communication partners, The step of determining the first communication partner from the simultaneous communication, and the transmission weight used when performing transmission weighting on the transmission signal in the first communication sequence of the first communication partner are set for the first communication partner. A step of determining the first transmission weight by linear calculation from the channel response matrix, a communication sequence other than the first communication sequence of the first communication partner, and the second communication In order to determine the transmission weight to be used when performing transmission weighting on the transmission signal in the hand communication sequence, the receiving antennas or reception beams of all the communication partners other than the communication partner whose transmission weight is to be determined are determined for each communication partner. And the channel response vector group between the transmission antennas of the transmission apparatus and the channel response vector between the reception beam formed at the first communication partner and the transmission antenna with respect to the first transmission weight. From the step of determining an interference space basis vector group, the step of calculating an orthogonal space channel response matrix that extracts components orthogonal to the interference space basis vector from the channel response matrix for all communication partners, and the orthogonal space channel response matrix, Determining a second transmission weight obtained by a linear operation; and an interference space basis vector. Unit vectors obtained by using the orthogonalization method for the row vector group of the complex conjugate matrix of the communication partner's channel response matrix, and the transmission side eigenvector obtained when singular value decomposition of the communication partner's channel response matrix Or calculating a vector obtained by QR decomposition, and an inner product value of the vectors related to different communication partners
P _{kl, ab} = (ω _{k, a} , ω _{l, b} )
(K ≠ _l, ω _{k, a} is the a th vector for the k th communication partner, and ω _{l, b} is the b th vector for the l th communication partner.)
And a vector is selected so that the inner product value becomes low, and an interference space basis vector is determined.

The present invention according to claim 3 includes a plurality of transmission antenna elements, and is provided between a reception antenna and a transmission antenna element included in a communication partner, or between a reception beam formed using the reception antenna and a transmission antenna. A wireless communication method for estimating a channel response matrix representing a propagation environment of the transmission, determining a transmission weight suitable for the propagation environment, performing transmission weighting on the transmission signal, and transmitting the transmission signal to a plurality of communication partners, If communication is performed at the same time, the step of determining the first communication partner and the transmission of the first communication sequence of the first communication partner are performed. Determining communication partners that do not have sufficient decoding devices, determining these communication partners as non-interference communication partners, and a channel response matrix for the non-interference communication partner And determining a first interference space basis vector and calculating a first orthogonal space channel response matrix obtained by extracting a component orthogonal to the first interference space basis vector from the channel response matrix for the first communication partner. Determining a first transmission weight of the first communication partner obtained by linear calculation from the first orthogonal spatial channel response matrix, and other than the first communication sequence of the first communication partner In order to determine the transmission weight used for the communication sequence and the communication sequence of the second communication partner, for each communication partner, the transmission antennas and the reception beams and transmissions of all the communication partners other than the communication partner whose transmission weight is to be determined are transmitted. A group of channel response vectors between the transmission antennas of the apparatus, and between the reception beam formed at the first communication partner and the transmission antenna with respect to the first transmission weight. A step of determining interference space basis vectors of all communication partners from the channel response vector, and an orthogonal spatial channel response matrix obtained by extracting components orthogonal to the interference space basis vectors from the channel response matrix for all communication partners. A step of determining a second transmission weight obtained by a linear operation from the orthogonal spatial channel response matrix, and orthogonal to a row vector group of a complex conjugate matrix of a channel response matrix of a communication partner as an interference space basis vector candidate Calculating the unit vector obtained by using the quantization method, the eigenvector on the transmission side obtained by singular value decomposition of the channel response matrix of the communication partner, or the vector obtained by QR decomposition, and related to different communication partners Inner product value of the vector
P _{kl, ab} = (ω _{k, a} , ω _{l, b} )
(K ≠ _l, ω _{k, a} is the a th vector for the k th communication partner, and ω _{l, b} is the b th vector for the l th communication partner.)
And a vector is selected so that the inner product value becomes low, and an interference space basis vector is determined.

The present invention according to claim 4 includes a plurality of transmission antenna elements, and is provided between a reception antenna and a transmission antenna element included in a communication partner, or between a reception beam formed using the reception antenna and a transmission antenna. A wireless communication method for estimating a channel response matrix representing a propagation environment of the transmission, determining a transmission weight suitable for the propagation environment, performing transmission weighting on the transmission signal, and transmitting the transmission signal to a plurality of communication partners, When performing a step of determining a plurality of first communication partners and performing transmission of the first communication sequence of the first communication partner during simultaneous communication, interference due to this communication sequence cannot be largely removed or removed. determines no communication partner sufficient decoding apparatus to, determining those communication partner as the non-interference communication partner among a plurality of the first communication partner, the interference The channel response matrix for non-communication partner communicatively partner, from the channel response matrix for the non-interference communication partner, and determining the first interference spatial basis vectors, from the channel response matrix for the first communication partner, the first A step of calculating a first orthogonal spatial channel response matrix obtained by extracting a component orthogonal to the interference space basis vector, and a first communication partner first obtained by linear calculation from the first orthogonal spatial channel response matrix A transmission weight is determined for each communication partner in order to determine a transmission weight used for the communication sequence other than the first communication sequence of the first communication partner and the communication sequence of the second communication partner. The channel response vector between the receiving antenna or receiving beam of all the communication partners other than the communication partner to be determined and the transmitting antenna of the transmitting device Determining a set of interference space basis vectors for all communication partners from a channel response vector between a reception beam formed at the first communication partner and a transmission antenna with respect to the first transmission weight, A step of calculating an orthogonal spatial channel response matrix in which a component orthogonal to the interference space basis vector is extracted from the channel response matrix of the communication partner, and a second transmission weight obtained by linear calculation from the orthogonal spatial channel response matrix The singular value decomposition of the unit vector obtained by using the orthogonalization method to the row vector group of the complex conjugate matrix of the communication partner's channel response matrix and the communication partner's channel response matrix as the interference space basis vector candidates Step for calculating the transmission side eigenvectors obtained by this or the vector obtained by QR decomposition. And the inner product value of the vectors related to different communication partners
P _{kl, ab} = (ω _{k, a} , ω _{l, b} )
(K ≠ _l, ω _{k, a} is the a th vector for the k th communication partner, and ω _{l, b} is the b th vector for the l th communication partner.)
And a vector is selected so that the inner product value becomes low, and an interference space basis vector is determined.

According to a fifth aspect of the present invention, in the wireless communication method according to the first to fourth aspects, an orthogonalization method is used as a row vector group of a complex conjugate matrix of a channel response matrix of a communication partner as an interference space basis vector. A step of calculating a unit vector obtained, a transmission-side eigenvector obtained by performing singular value decomposition on a channel response matrix of a communication partner , or a vector obtained by QR decomposition , is there.

According to a sixth aspect of the present invention, in the wireless communication method according to the first to fourth aspects, an orthogonalization method is used for a row vector group of a complex conjugate matrix of a channel response matrix of a communication partner as an interference space basis vector candidate. The step of calculating the unit vector obtained in step 1, the eigenvector on the transmission side obtained when performing singular value decomposition on the channel response matrix of the communication partner, or the vector obtained by QR decomposition, and the obtained vector as the transmission weight Estimating the signal-to-noise ratio of the communication sequence obtained in this way, calculating the inner product value of the vectors related to different communication partners, and calculating the signal pair when those vectors are used from the correlation between the vectors. Estimating the degradation of the noise ratio, and selecting a combination with the highest expected transmission capacity when performing communication. It is a wireless communication method according to symptoms.

According to a seventh aspect of the present invention, in the wireless communication method according to the first to fourth aspects, the orthogonalization method is used as a row vector group of a complex conjugate matrix of a channel response matrix of a communication partner as an interference space basis vector candidate. The step of calculating the unit vector obtained in step 1, the eigenvector on the transmission side obtained when performing singular value decomposition on the channel response matrix of the communication partner, or the vector obtained by QR decomposition, and the obtained vector as the transmission weight A signal when using these vectors from the step of estimating the signal-to-noise ratio of the communication sequence originally obtained, the step of calculating the inner product value of the vectors related to different communication partners, and the correlation between the vectors Estimate the degradation of the noise-to-noise ratio and select a combination that can achieve the maximum transmission rate with acceptable quality in view of the selectable modulation method That is a wireless communication method characterized by comprising the steps.

The present invention according to claim 8 is the wireless communication method according to any one of claims 1 to 4, wherein after determining the transmission weight or the first and second transmission weights, the interference space basis vector or the first and second transmission weights are determined again. Calculating a channel response vector between the reception beam of the communication partner corresponding to the determined transmission weight of the communication partner and the transmission antenna element as the interference space basis vector of the communication partner, an orthogonal spatial channel response matrix, or 1. calculating a second orthogonal spatial channel response matrix based on a newly defined interference space basis vector or using the first and second interference space basis vectors, and the obtained orthogonal spatial channel response matrix Alternatively, a new transmission weight or first and second transmission weights are determined from the first and second orthogonal spatial channel response matrices. And a step that is these steps a wireless communication method characterized by repeated any number of times.

The present invention according to claim 9 includes a transmitting station having a plurality of antenna elements and a plurality of communication counterpart stations having one or a plurality of antenna elements. Via a MIMO (Multiple Input Multiple Output) channel composed of antenna elements or beams formed on these antenna elements, one or a plurality of signal sequences are sent to the same frequency channel and the same time for a plurality of communication counterpart stations. In a wireless communication system capable of performing MIMO communication by spatial multiplexing, in a wireless communication apparatus that transmits a signal by means of L (1) to L (Mu) spatial multiplexing for Mu communication partners, transmission spatial multiplexing is performed. MT (MT> 1: adjustment) which is equal to or greater than several Mu × (L (1) + L (2) +... + L (Mu)) ) It has antenna elements and is connected to each antenna element. When receiving, it converts the received signal into a baseband signal and outputs it to the channel information acquisition circuit. When transmitting, it transmits the transmission signal as a radio signal from the antenna element. And the channel response matrix for the communication partner is estimated from the signal input from the radio unit, and when the communication partner and the transmission signal are determined, the channel response matrix of the communication partner for transmission is output to the interference space arithmetic circuit Channel information acquisition circuit, an interference space calculation circuit for calculating an interference space basis vector for each communication partner based on a channel response matrix input from the channel information acquisition circuit, and an interference space input from the interference space calculation circuit Based on the basis vector and the channel response matrix input from the channel response matrix acquisition circuit, An orthogonal spatial computation circuit that calculates an orthogonal spatial channel response matrix that reduces external interference, and an orthogonal spatial channel response matrix calculated by the orthogonal spatial computation circuit as an input signal, and performing linear computation on the orthogonal spatial channel response matrix A transmission weight calculation circuit for determining the transmission weight obtained in step, outputting the transmission weight to the transmission signal conversion circuit, and outputting the transmission mode comprising the modulation scheme and coding rate applied to each communication sequence to the modulation circuit and the data division circuit, and A transmission weight determination block including a channel response matrix acquisition circuit, an interference space calculation circuit, an orthogonal space calculation circuit, and a transmission weight calculation circuit, a data division circuit that divides transmission data into the number of communication sequences according to the transmission mode, and A modulation circuit that modulates the signal sequence, and a transmission weight determined by the transmission weight calculation circuit for each modulated communication sequence Then, use the orthogonalization method for the row vector group of the complex conjugate matrix of the channel response matrix of the communication partner as a candidate for the interference space basis vector, and the transmission signal conversion circuit that outputs to the radio unit connected to the corresponding antenna element Of the unit vector obtained in step S1, the transmission side eigenvector obtained when singular value decomposition of the channel response matrix of the communication partner, or the vector obtained by QR decomposition, and the vector related to a different communication partner Inner product value
P _{kl, ab} = (ω _{k, a} , ω _{l, b} )
(K ≠ _l, ω _{k, a} is the a th vector for the k th communication partner, and ω _{l, b} is the b th vector for the l th communication partner.)
And a determination unit that selects a vector so that the inner product value becomes low and determines an interference space basis vector.

The present invention according to claim 10 is the wireless communication apparatus according to claim 9 , wherein the transmission weight determination block estimates a channel response matrix from a signal input from a wireless unit, and a communication partner and a transmission signal are After the determination, the channel information acquisition circuit that outputs the channel response matrix of the communication partner to the interference space candidate calculation circuit and the interference space basis vector candidate that is the candidate of the interference space basis vector from the input channel response matrix, Unit vector obtained by using orthogonalization method for row vector group of complex conjugate matrix of channel response matrix, transmission side eigenvector obtained when singular value decomposition of channel response matrix of communication partner, or this eigenvector can be approximated Interference space candidate calculation circuit that outputs the vector obtained by linear calculation to the communication space selection circuit and the actual transmission Communication space selection circuit for selecting an interference space basis vector used as a communication partner, the number of communication sequences, and an interference space, an interference space basis vector input from the communication space selection circuit, and a channel response input from the channel information acquisition circuit An orthogonal spatial computation circuit that computes an orthogonal spatial channel response matrix that reduces interference with a non-communication partner, and an orthogonal spatial channel response matrix computed in the orthogonal spatial computation circuit as an input signal, A transmission weight calculation circuit that determines a transmission weight obtained by performing a linear operation on a response matrix, outputs the transmission weight to a transmission signal conversion circuit, and outputs a transmission mode to be applied to each communication series to a modulation circuit and a data division circuit. A wireless communication device characterized by the above.

The present invention according to claim 11 is the wireless communication apparatus according to claim 9 , wherein the transmission weight determination block estimates a channel response matrix from a signal input from a wireless unit, and a communication partner and a transmission signal are After the determination, the channel information acquisition circuit that outputs the channel response matrix of the communication partner to the interference space candidate calculation circuit and the interference space basis vector candidate that is the candidate of the interference space basis vector from the input channel response matrix, Unit vector obtained by using orthogonalization method for row vector group of complex conjugate matrix of channel response matrix, transmission side eigenvector obtained when singular value decomposition of channel response matrix of communication partner, or this eigenvector can be approximated Interference space candidate calculation circuit that outputs the vector obtained by linear calculation to the communication space selection circuit and the actual transmission Select a communication partner, the number of communication sequences, and an interference space basis vector to be used as an interference space, and further select a quasi-interference space basis vector that is not completely orthogonal to the channel response matrix from among the interference space basis vector candidates that were not selected. Communication space selection circuit that outputs to the orthogonal space arithmetic circuit, interference space basis vectors and quasi-interference space basis vectors input from the communication space selection circuit, and a channel response matrix input from the channel information acquisition circuit , An orthogonal spatial computation circuit that calculates an orthogonal spatial channel response matrix that reduces interference with non-communication partners, and an orthogonal spatial channel response matrix calculated by the orthogonal spatial computation circuit as an input signal, Determine the transmission weight obtained by performing linear operation and output it to the transmission signal conversion circuit. It is a wireless communication apparatus characterized by comprising a transmission weight calculation circuit which outputs a transmission mode to be applied to the modulation circuit and the data dividing circuit.

The present invention according to claim 12 is the wireless communication apparatus according to claim 9 , wherein the transmission weight determination block estimates a channel response matrix from a signal input from a wireless unit, and a communication partner and a transmission signal are After the determination, the channel information acquisition circuit that outputs the channel response matrix of the communication partner to the interference space candidate calculation circuit and the interference space basis vector candidate that is the candidate of the interference space basis vector from the input channel response matrix, Unit vector obtained by using orthogonalization method for row vector group of complex conjugate matrix of channel response matrix, transmission side eigenvector obtained when singular value decomposition of channel response matrix of communication partner, or this eigenvector can be approximated Interference space candidate calculation circuit that outputs vector obtained by linear calculation to communication space selection circuit, and the influence of time variation , By an error of the apparatus, the vector is expected to be an interference space when actually transmitting, the additional interference space selection circuit for outputting as the additional interference space basis vectors, actual communication partner performs transmission, communication Select the number of sequences and the interference space basis vector to be used as the interference space, and then select and add a quasi-interference space basis vector that is not completely orthogonal to the channel response matrix from the interference space basis vector candidates that were not selected. A communication space that converts the additional interference space basis vector input from the interference space selection circuit to a vector orthogonal to the interference space basis vector, and outputs it to the orthogonal space arithmetic circuit in addition to the interference space basis vector or the quasi-interference space basis vector. Selection circuit, interference space basis vectors, quasi-interference space basis vectors input from the communication space selection circuit, and channel information. An orthogonal spatial arithmetic circuit that calculates an orthogonal spatial channel response matrix that reduces interference with other parties that communicate with the channel response matrix input from the acquisition circuit, and an orthogonal spatial channel response matrix that is calculated in the orthogonal spatial arithmetic circuit Is used as an input signal, the transmission weight obtained by performing a linear operation on the orthogonal spatial channel response matrix is determined, output to the transmission signal conversion circuit, and the transmission mode applied to each communication sequence is output to the modulation circuit and data division circuit And a second transmission weight calculation circuit.

According to a thirteenth aspect of the present invention, in the wireless communication apparatus according to the ninth aspect , the transmission weight determining block performs transmission preferentially from among the communication partners when the communication partner and the transmission signal are determined. A transmission rank setting circuit that determines the number of communication partners and the number of first communication series, and outputs information of the first communication partner and other communication partners to a channel information acquisition circuit, and a channel from a signal input from the radio unit When a response matrix is estimated and a communication partner and a transmission signal are input from the transmission rank setting circuit, the channel response matrix of the first communication partner and the number of first communication sequences are sent to the first transmission weight calculation circuit. A channel information acquisition circuit that outputs the channel response matrix of all communication partners to the second interference space candidate calculation circuit, and a first transmission weight by linear calculation from the input channel response matrix of the first communication partner Determine the specified number and use the first transmission weight calculation circuit that outputs to the interference space candidate calculation circuit and the channel response matrix input from the channel information acquisition circuit, and orthogonal to the row vector group of the complex conjugate matrix of the channel response matrix A communication space selection circuit can be used to generate a unit vector obtained by using the conversion method, a transmission-side eigenvector obtained by singular value decomposition of the channel response matrix of the communication partner, or a vector obtained by linear operation that can be approximated to this eigenvector. The second interference space candidate calculation circuit that outputs the signal, and a vector that is expected to become an interference space when actually transmitting due to the influence of time variation and device error is output as an additional interference space basis vector Additional interference space selection circuit that performs transmission, the communication partner that actually performs transmission, the number of communication sequences, and the interference space basis vector used as the interference space Select, further from the selected not interference spatial basis vectors candidate channel select quasi interference space basis vectors does not perpendicular completely the response matrix, additional interference space added interference is input from the selection circuit space basis vectors A second communication space selection circuit that converts the vector into a vector orthogonal to the interference space basis vector and outputs the vector to the second orthogonal space arithmetic circuit in addition to the interference space basis vector or the quasi-interference space basis vector; Orthogonal spatial channel response that reduces interference to other parties that are communicating from the interference space basis vectors and quasi-interference space basis vectors input from the communication space selection circuit and the channel response matrix input from the channel information acquisition circuit Input a second orthogonal spatial arithmetic circuit for calculating a matrix and an orthogonal spatial channel response matrix calculated in the second orthogonal spatial arithmetic circuit The transmission weight obtained by performing a linear operation on the orthogonal spatial channel response matrix is determined as a signal, output to the second transmission signal conversion circuit, and the transmission mode applied to each communication sequence is set to the modulation circuit and the data division circuit. A wireless communication apparatus comprising a second transmission weight calculation circuit for outputting.

The present invention according to claim 14 is the wireless communication apparatus according to claim 9 , wherein the transmission weight determination block has a plurality of transmissions preferentially performed from among the communication partners when the communication partner and the transmission signal are determined. A transmission rank setting circuit for determining the number of first communication partners and the number of first communication series, and outputting information of the first communication partner and other communication partners to the channel information acquisition circuit, and a signal input from the radio unit When a communication partner and a transmission signal are input from the transmission rank setting circuit, the channel response matrix of the first communication partner and the number of first communication sequences are calculated from the first orthogonal space arithmetic circuit. And a channel information acquisition circuit that outputs the channel response matrix of all communication partners to the second interference space candidate calculation circuit, and the input channel response matrix of the first communication partner. ,each A first interference space calculation circuit that determines a first interference space basis vector of one communication partner and outputs the first interference space basis vector to the first orthogonal space calculation circuit; and a first interference input from the first interference space calculation circuit A first number is determined for the first transmission weight by linear calculation from the spatial basis vector and the channel response matrix of the first communication partner input from the channel information acquisition circuit, and is output to the interference space candidate calculation circuit. Using the channel response matrix input from the transmission weight calculation circuit and the channel information acquisition circuit, the unit vector obtained by using the orthogonalization method for the row vector group of the complex conjugate matrix of the channel response matrix and the channel of the communication partner A communication space selection circuit is used for the transmission side eigenvector obtained when singular value decomposition of the response matrix or a vector obtained by linear operation that can be approximated with this eigenvector. A second interference spatial candidate calculation circuit for outputting, influence of fluctuation time, the error of the device, a vector is expected to be an interference space when actually transmitting, the output as an additional interference space basis vectors A channel response matrix that selects an additional interference space selection circuit to be performed, a communication partner actually performing transmission, the number of communication sequences, and an interference space basis vector to be used as an interference space, and further selected from interference space basis vector candidates not selected Selects a quasi-interference space basis vector that is not completely orthogonal, converts the additional interference space basis vector input from the additional interference space selection circuit into a vector orthogonal to the interference space basis vector, In addition to the quasi-interference space basis vector, the second communication space selection circuit that outputs to the second orthogonal space arithmetic circuit and the second communication space selection circuit A second method for calculating an orthogonal spatial channel response matrix that reduces interference with other parties that communicate with each other based on the input interference space basis vector and quasi-interference space basis vector and the channel response matrix input from the channel information acquisition circuit. The orthogonal spatial arithmetic circuit and the orthogonal spatial channel response matrix calculated by the second orthogonal spatial arithmetic circuit are used as input signals, and a transmission weight obtained by performing linear arithmetic on the orthogonal spatial channel response matrix is determined, and the transmission signal is determined. A radio communication apparatus comprising a second transmission weight calculation circuit that outputs to a conversion circuit and outputs a transmission mode to be applied to each communication series to a modulation circuit and a data division circuit.

The present invention according to claim 15 is the wireless communication apparatus according to claim 9 , wherein the transmission weight determination block has a plurality of transmissions preferentially performed from among the communication partners when the communication partner and the transmission signal are determined. A transmission rank setting circuit for determining the number of first communication partners and the number of first communication series, and outputting information of the first communication partner and other communication partners to the channel information acquisition circuit, and a signal input from the radio unit When a communication partner and a transmission signal are input from the transmission rank setting circuit, the channel response matrix of the first communication partner and the number of first communication sequences are calculated from the first orthogonal space arithmetic circuit. And a channel information acquisition circuit that outputs the channel response matrix of all communication partners to the second interference space candidate calculation circuit, and the input channel response matrix of the first communication partner From A first interference space candidate vector circuit that calculates a first interference space basis vector candidate of each first communication partner and outputs it to the first communication space operator circuit, and an input first interference space basis vector candidate The first interference space basis vector to be used as the first interference space is selected from the first interference space basis vector candidates that are not selected, and the channel response matrix is not completely orthogonal to the first one. A first communication space selection circuit that selects a quasi-interference space basis vector and outputs it to the first orthogonal space operation circuit; a first interference space basis vector input from the first communication space operation circuit; A prescribed number of first transmission weights are determined by linear calculation from the quasi-interference space basis vector and the channel response matrix of the first communication partner input from the channel information acquisition circuit, and second interference space candidate calculation is performed. A unit obtained by using an orthogonalization method for the row vector group of the complex conjugate matrix of the channel response matrix using the first transmission weight calculation circuit output to the path and the channel response matrix input from the channel information acquisition circuit A second interference space candidate that outputs to the communication space selection circuit a vector, a transmission-side eigenvector obtained by singular value decomposition of the channel response matrix of the communication partner, or a vector obtained by linear operation that can be approximated to this eigenvector An additional interference space selection circuit that outputs, as an additional interference space basis vector, a vector that is expected to be an interference space when actually transmitting due to the influence of time variation and device errors, Select the communication partner that actually performs transmission, the number of communication sequences, and the interference space basis vector used as the interference space. A quasi-interference space basis vector that is not completely orthogonal to the channel response matrix is selected from the interference space basis vector candidates, and the additional interference space basis vector input from the additional interference space selection circuit is orthogonal to the interference space basis vector. In addition to the interference space basis vector or the quasi-interference space basis vector, the second communication space selection circuit that outputs to the orthogonal space arithmetic circuit, the interference space basis vector input from the communication space selection circuit, In an orthogonal space arithmetic circuit that calculates an orthogonal spatial channel response matrix that reduces interference with other parties that perform communication from an interference space basis vector and a channel response matrix input from a channel information acquisition circuit, and an orthogonal space arithmetic circuit By using the calculated orthogonal spatial channel response matrix as the input signal and performing a linear operation on the orthogonal spatial channel response matrix A transmission weight calculating circuit that determines a transmission weight to be transmitted, outputs the transmission weight to a transmission signal conversion circuit, and outputs a transmission mode to be applied to each communication series to a modulation circuit and a data division circuit. is there.

The present invention according to claim 16 is the wireless communication apparatus according to claim 9 , wherein the transmission weight determination block is configured to transmit a plurality of transmissions preferentially from the communication partners when a communication partner and a transmission signal are determined. A transmission rank setting circuit for determining the number of first communication partners and the number of first communication series, and outputting information of the first communication partner and other communication partners to the channel information acquisition circuit, and a signal input from the radio unit When a communication partner and a transmission signal are input from the transmission rank setting circuit, the channel response matrix of the first communication partner and the number of first communication sequences are calculated from the first orthogonal space arithmetic circuit. Output to the first interference space candidate calculation circuit, and the channel response matrix of all communication partners is output to the second interference space candidate calculation circuit, so that it is difficult to completely eliminate the interference due to the first communication series. Communication partner From the channel information acquisition circuit that outputs at least part of the channel response matrix to the first additional interference space selection circuit and the input channel response matrix of the first communication partner, the first interference of each first communication partner A first interference space candidate calculation circuit that calculates a spatial basis vector candidate and outputs it to the first communication space calculation circuit, and becomes an interference space during actual transmission due to the influence of time variation and device errors. A first additional interference space selection circuit that outputs, as an additional interference space basis vector, an interference space basis vector obtained from a channel response matrix of a communication partner with which communication is impossible The first interference space basis vector candidate to be used as the first interference space is selected from the first interference space basis vector candidates, and the first interference space basis vector candidates not selected are selected. From within, it selects the first quasi-interference space basis vectors does not completely is orthogonal to the channel response matrix, the first additional interference space basis vectors input from the first additional interference spatial selection circuit, the first A first communication space selection circuit for converting to a vector orthogonal to the interference space basis vector and outputting to the first orthogonal space arithmetic circuit in addition to the first interference space basis vector or the first quasi-interference space basis vector; the first interference space basis vectors and the first quasi-interference spatial basis vectors input from the first communication space arithmetic circuit, the channel response matrix of the first communication partner is input from the channel information acquiring circuit, linear A predetermined number of first transmission weights are determined by calculation, and the first transmission weight calculation circuit that outputs the first transmission weight to the second interference space candidate calculation circuit, and the channel response input from the channel information acquisition circuit. From the answer matrix and the first transmission weight input from the first transmission weight calculation circuit, a second interference space basis vector candidate that is a second interference space basis vector candidate is output to the communication space selection circuit. Interference space candidate calculation circuit and second addition that outputs a vector that is expected to become an interference space when actually transmitting due to the influence of time variation or device error as an additional interference space basis vector The channel response matrix is selected from the interference space selection circuit, the communication partner that actually performs transmission, the number of communication sequences, and the interference space basis vector to be used as the interference space. fully select quasi interference space basis vectors does not orthogonal, the additional interference space basis vectors input from the additional interference space selection circuit, the interference spatial basis vectors In addition to the interference space basis vector or quasi-interference space basis vector, the communication space selection circuit that outputs to the orthogonal space operation circuit, the interference space basis vector input from the communication space selection circuit, and the quasi-interference space An orthogonal space arithmetic circuit that calculates an orthogonal spatial channel response matrix that reduces interference with a party other than the communication partner from the basis vector and the channel response matrix input from the channel information acquisition circuit, and an orthogonal space arithmetic circuit Using the orthogonal spatial channel response matrix as an input signal, the transmission weight obtained by performing a linear operation on the orthogonal spatial channel response matrix is determined, output to the transmission signal conversion circuit, and the transmission mode applied to each communication sequence is modulated by the modulation circuit And a transmission weight calculation circuit for outputting to the data division circuit.

  According to the present invention, when the communication partner and the data to be transmitted are determined, the channel response matrix of the corresponding communication partner is sent to the orthogonal space arithmetic circuit and the interference space arithmetic circuit. The vector group is calculated from the channel response matrix, the orthogonal spatial channel response matrix is calculated from the channel response matrix and the interference space basis vector group, output to the transmission weight calculation circuit, and each signal for each communication partner is calculated by the transmission weight calculation circuit. The transmission weight at each antenna of the series is calculated. Thereby, the interference signal to those other than the desired communication partner is reduced.

Hereinafter, various embodiments of the present invention will be described with reference to the drawings.
(First embodiment)
FIG. 1 shows a configuration example of a transmission unit in the first embodiment of the present invention. In Figure 1, 100 is data dividing circuit, 101-1 to 101-L is the modulation circuit, 102 transmitting signal conversion circuit, 103-1 to _{103-M T} wireless unit, 104-1 to _{104-M T} antenna , 105 is a transmission weight calculation circuit, 106 is an orthogonal space calculation circuit, 108 is an interference space calculation circuit, 107 is a channel response matrix acquisition circuit, and 110 is a transmission weight determination block.

Here, M _{T is} the number of antenna elements of the transmitting device, M _{R is} the number of all receiving antenna elements of all communicating parties that perform simultaneous communication, M _R (k) is the number of receiving elements of the kth communicating party, and the same frequency. Let L be the total number of communication sequences for all communication partners transmitting in the band.

As in the prior art, the antenna 104-1 to _{104-M T} and the radio unit 103-1 to _{103-M T} is capable of transmitting and receiving radio signals, the radio unit 103-1～103- based on the converted received signal in M _T, the channel between the channel information obtaining circuit 109, a receiving antenna or reception beams, each antenna 104-1 to 104-M _T and a plurality of communication partner of the transmission unit A response matrix can be estimated. Although this method obtains the channel response matrix is not specified here, the feedback included in the estimate based on whether or received signal information obtained when performing the reception of the known signal in the antenna 104-1 to 104-M _T Information on the channel response matrix for each communication partner is acquired by the information included in the information.

  When the communication partner and the data to be transmitted are determined, the channel response matrix of the corresponding communication partner is output to the orthogonal space arithmetic circuit 106 and the interference space arithmetic circuit 107. In the interference space calculation circuit 107, an interference space basis vector group is calculated from the channel response matrix and output to the orthogonal space calculation circuit 106.

  The orthogonal space arithmetic circuit 106 calculates an orthogonal spatial channel response matrix from the channel response matrix and the interference space basis vector group and outputs the transmission weight arithmetic circuit 105.

  The transmission weight calculation circuit 105 calculates the transmission weight at each antenna of each signal series for each communication partner.

Transmission data is divided into L signal series by data division circuit 100 and input to modulation circuits 101-1 to 101-L. Here, after a preamble signal or the like for MIMO channel estimation is added and modulated, these signals are input to the transmission signal conversion circuit 102. Here, the transmission data is multiplied by a transmission weight, but is input to the radio unit 103-1 to _{103-M T,} it is transmitted as a radio signal through the antenna 104-1 to _{104-M T.}

An example of the control method is shown below. First, the channel information acquisition circuit 108 includes an all-space channel response matrix H _all (M _R × M _T matrix),
H _all = [H ₍₁₎ ^T _, H ₍₂₎ ^T _, ..., H _(Mu) ^T ] ^T
Estimate The total spatial channel matrix obtained here is output to interference space arithmetic circuit 107 and orthogonal spatial arithmetic circuit 106.

  The interference space calculation circuit 107 uses, for each communication partner, a channel response vector formed between a reception antenna or a reception beam other than the communication partner and a transmission antenna as an interference space.

The matrix _{G (k) (M R '} (k) × M T matrix), and an interference spatial channel response matrix consists response matrix H _(k) other than the channel response matrix _{H (l) (l ≠ k} ), _Let M _R ′ (k) = M _R −M _R (k). Therefore, the interference spatial channel response matrix G ( _k ) is expressed as follows.

G ( _k ) = [H ( ₁ ) ^T , ..., H ( _k - ₁ ) ^T , H ( _k + ₁ ) ^T , ..., H ( _Mu ) ^T ] ^T

When a signal is transmitted to the space constituted by this interference spatial channel response matrix G ( _k ), it becomes an interference signal to other communication partners. The a-th row vector of the interference spatial channel response matrix _{G (k) g k,} is _a, the interference spatial basis vectors group _e k row components of the channel response matrix _{G (k), 1, e} k, 2, ·· .., E _k , _MR ′ ( _k ) are calculated as follows by using the orthogonalization method to calculate the interference space basis vector group.

E _(k) = [e _{k, 1} ^T , e _{k, 2} ^T , ... e _{k, MR} ' _(k) ^T ] ^T is an interference space basis vector group for the k th communication partner. (A, b) represents an inner product value of the vector a and the vector b, and (a, b) = a · b ^H / (| a | · | b |).

Further, as this interference space basis vector, a channel response matrix H _(l) (l ≠ k) corresponding to a communication partner other than the communication partner is subjected to singular value decomposition, and H _(l) = U _(l) D _(l) It is also possible to select a conjugate complex vector of V _(l) ^{H on} the transmission side eigenvector V _(l) = (v _{l, 1} ,..., V _{l, MR (l)} ).

Next, the orthogonal space arithmetic circuit 106 obtains an orthogonal space channel response matrix H ′ _(k) in the kth communication partner that has no interference with the other communication partner other than the kth or has a small interference. This orthogonal spatial channel response matrix H ′ _(k) is a row vector of the communication partner H _(l) (l ≠ k) that performs simultaneous communication other than the k-th communication partner of the orthogonal spatial channel response matrix H _(k). It is a matrix composed only of completely orthogonal vectors. If the transmission weight is determined in this orthogonal space, interference with other communication partners can be eliminated.

The orthogonal spatial channel response matrix H ′ _(k) is the channel response matrix H ( _k )
H ( _k ) = [H _{k, 1} ^T , h _{k, 2} ^T ,..., H _{k, M} R _(k) ^T ] ^T
And interference space basis vector group E _(k)
E _(k) = [e _{k, 1} ^T , e _{k, 2} ^T , ... e _{k, MR} ' _(k) ^T ] ^T
And can be expressed as follows.

Here, (a, b) represents the inner product value of the vectors a and b, h _{k, j} represents the j th row vector of the channel response matrix H _(k) for the k th communication partner, and h ′ _{k , J} represents the j th row vector of the orthogonal spatial channel response matrix H ′ _(k) for the k th communication partner. In this way, the orthogonal spatial channel response matrix H ′ _(k) = [h ′ _{k, 1} ^T , h ′ _{k, 2} ^T , ... H ′ _{k, MR (k)} ^T ] ^T is obtained.

'Upon determining the transmission weight by using the information of _(k), orthogonal space channel response matrix H' thus determined orthogonal space channel response matrix H _(k) is to the other communication party, completely orthogonal Therefore, an interference signal is not given to other communication partners.

For example orthogonalization method using a column vector of 'conjugate complex transposed matrix H _{(k)' (k)} ^H orthogonal spatial channel response matrix H, that obtained base vector to the transmission weight, interference with other communication partners The transmission weight can be determined so that no signal is generated. QR decomposition is performed on the conjugate complex transpose matrix H ′ _(k) ^H ,
H ′ _(k) ^H = Q ′ _(k) R ′ _(k)
The same applies to Q ′ _(k) obtained as a transmission weight.

The transmission weight thus obtained is W _(k) (M _T × L (k) matrix), and the total transmission weight W _all used for all communication partners is:
W _all = [W ₍₁₎ , W ₍₂₎ ,..., W _(k) ]
When communication is performed using the transmission weight thus obtained, the reception signal Y _all received at each communication partner can be expressed as follows.

Here, Z _all is a thermal noise vector corresponding to the received signals of all communication partners. If attention is paid only to the communication partner 1, the received signal Y ₍₁₎ can be expressed as follows.

Here, since the transmission weights W _{(2) to} W _(Mu) are vector groups that are completely orthogonal to the channel response matrix of the communication counterpart 1, the above equation can be rewritten as follows.

  Therefore, it can be handled in the same way as communication with a single communication partner.

Here, the transmission weight W _(k) is a matrix obtained by linear operation using orthogonal spatial channel response matrix H _'(k).

For example, singular value decomposition H ′ _(k) = U ′ _(k) D ′ _(k) V ′ _(k) ^H as transmission weight W _(k)
Eigenvector V ′ _(k)
V ′ _(k) = (v ′ _{k, 1} ,..., V ′ _{k, MR (k)} )
Is selected, it is possible to obtain a high transmission capacity corresponding to the eigenvalue D ′ _(k) .

In addition, since the reception power of each communication sequence at the communication partner can be estimated from the eigenvalues λ ′ _{k, 1 to} λ ′ _{k, L} which are diagonal components of the eigenvalue D ′ _(k) , The determination of L _(k) and the transmission mode consisting of the modulation scheme and coding rate can be set based on the value of this eigenvalue.

  For example, a table of transmission modes corresponding to the received power is created in advance. If the signal to noise ratio (SNR) is 5 to 10 dB, BPSK (Binary Phase Shift Keying) is used. If it is 10 to 16 dB, QPSK (Quadrature Phase Shifting) is used. ), 16-value QAM (Quadrature Amplitude Modulation) for 16 to 23 dB, and 64-value QAM for 23 to 30 dB, and 256-value QAM for 30 dB or more can be used to determine the modulation method.

In addition, the number of communication sequences L (j) used for each communication partner can be selected smaller than the number of receiving antenna elements M _R (j) of the communication partner. In this case, the interference spatial channel response matrix G _(k) for obtaining the orthogonal spatial channel response matrix can be selected smaller, and can be formed as follows.
G _(k) = [(F ₍₁₎ ^H H ₍₁₎ ) ^T , ,, (F _(k-1) ^H H _(k-1) ) ^T , (F _{(k + 1)} ^H H _{(k + 1)} ) ^T , ..., (F _(Mu) ^H H _(Mu) ) ^T ] ^T

Here, F _(j) represents an initial weight matrix (M _R (j) × L ′ _k (j) matrix), and L ′ _k (j) represents the jth (j ≠ k) of the kth communication partner. This is the number of vectors selected as the interference space from the communication partners, and the column vectors of F _(j) are selected so as to be completely orthogonal.

As a simple example, when antenna selection is performed using an initial weight matrix, in each column vector of the initial weight matrix F _(k) , an element corresponding to the antenna to be selected is a real number, and other values are set to 0. Well, any number of receive antennas can be considered.

The vector number L ′ _k (j) is made equal to the vector number M _R (j). L ′ _k (j) = M _R (j)
By doing so, the first embodiment of the present invention is obtained.

If the vector number L ′ _k (j) is selected to be smaller than the vector prime M _R (j), the k-th communication partner has a larger transmission capacity instead of reducing the degree of freedom of the j-th communication partner. Can be built.

As the initial weight matrix F _(k) , for example, an orthogonalization method is used for the column vector of the channel response matrix H _(k) or the receiving side correlation matrix H _(k) H _(k) ^H of the communication partner, and the obtained base vector is Can be used. QR decomposition of H _(k)
H _(k) = Q _(k) R _(k)
The same applies to the obtained Q _(k) as the initial weight. Alternatively, the eigenvectors of H _(b) H _(h) ^H can be used, and the corresponding eigenvalues can be selected.

  FIG. 2 shows a transmission flow in the first embodiment of the present invention. In FIG. 2, before performing communication, a channel response matrix corresponding to a communication partner performing transmission is estimated in advance (step S101).

When the communication partner and transmission data are determined (step S102), an interference space basis vector is calculated for each communication partner using the channel response matrix of the corresponding communication partner (step S103), and the obtained interference space is obtained. An orthogonal spatial channel response matrix is calculated using the basis vectors and the channel response matrix estimated in step S101 (step S104). Using the orthogonal spatial channel response matrix determined for each communication partner, the transmission weight and the transmission mode to be applied are determined (step S105), the transmission data is input (step S106), and the data to be transmitted corresponds to the transmission mode. The data is divided into L series by the bit length (step S107), modulation processing and addition of a known signal are performed (step S108), and the processing in the signal conversion circuit is performed in symbol units, for example, a symbol for the kth communication partner. If the signal vector at S _k is S _k and the transmission weight for the k th communication partner is W _k , a signal conversion process for performing a signal conversion process of S _k → W _k · S _k is performed (step S109). In this way, the processed baseband signal is converted into an RF signal by a radio unit corresponding to each antenna and transmitted (step S110).

(Second Embodiment)
FIG. 3 shows a second embodiment of the present invention. In FIG. 3, reference numeral 205 denotes a transmission weight calculation circuit, 206 denotes an orthogonal space calculation circuit, 207 denotes a communication space selection circuit, 208 denotes a channel information acquisition circuit, and 209 denotes an interference space candidate calculation circuit. The communication space selection circuit 207 selects the number of communication partners and the number of spatial multiplexing for downlink.

As shown in FIG. 1, in the first embodiment, when a communication partner and data to be transmitted are determined, the interference space calculation circuit 107 calculates an interference space basis vector group from the channel response matrix. When the formed interference space basis vector group has a high correlation with the channel response matrix H _{(k) of the} communication partner, the absolute value of the orthogonal spatial channel response matrix obtained by the orthogonal space arithmetic circuit 106 becomes small, and the transmission capacity Is considered to be reduced. Therefore, the transmission capacity that can be transmitted can be increased by selecting the interference space basis vector before calculating the orthogonal spatial channel response matrix.

In the second embodiment, first, in the interference space candidate calculation circuit 209, the channel response matrix of each communication partner is used, and the row vector of the complex conjugate matrix of the channel response matrix H _(k) is obtained using the orthogonalization method. Interference space basis vector candidate Ω _(k)
Ω _(k) = [ω _{k, 1} ^T , ω _{k, 2} ^T , ... ω _{k, MR (k)} ^T ] ^T
Is calculated.

Here, the interference space basis vector candidate Ω (k) is calculated using the channel response matrix of the communication partner itself, unlike the interference space basis vector formed from the channel response matrix of the communication partner other than the communication partner. Is done. The interference space basis vector candidates ωk, _{1 to} ωk, _{1MR (k)} are vectors obtained by linear calculation from the channel response matrix H _(k) .

For example, singular value decomposition of the channel response matrix H _(k)
H _(k) = U _(k) D _(k) V _(k) ^H
Eigenvector V _(k)
V _(k) = (vk _{, 1} , ..., vk _{, MR (k)} )
Can be selected as ω _{k, j} .

The interference space basis vector candidate Ω _(k) is output from the interference space candidate calculation circuit 209 to the communication space selection circuit 207. In the communication space selection circuit 207, the input interference space basis vector candidate Ω _(k) is evaluated for the inner product values P _{kl and ab} between different communication partners.
P _{kl, ab} = (ω _{k, a} , ω _{1, b} ).

Here, k ≠ l, ω _{k, a} is the a th interference space basis vector for the k th communication partner, and ω _{l, b} is the b th interference space basis vector for the l th communication partner.

An example of selection is shown in FIG. FIG. 4 shows a case where there are three communication partners A to C.
M _R (A) = 4,
M _R (B) = 3,
M _R (C) = 3
It is assumed that the absolute value of the inner product value of the interference space basis vector candidate is shown. Here, it is possible to reduce the number of communication sequences and the number of vectors to be selected as the interference space on the assumption that the interference space basis vector candidates that increase the absolute value of the inner product value are not used for other communication partners.

  In FIG. 4, the second interference space basis vector candidate of the communication partner C can be removed because the absolute value of the inner product value for the other communication partner is high. Or, when it is desired to increase the transmission capacity to the communication partner C with priority, the second eigenvector of the communication partner A and the first eigenvector of the communication partner B have a high correlation with the vector corresponding to the communication partner C. The interference space basis vector candidates corresponding to these can not be selected.

At this time, when the communication quality is selected with priority, and the absolute value of the inner product value is high, the one with the lower expected communication quality can be selected. As the communication quality, the communication quality that can be expected when the interference space basis vector candidate is used as a transmission weight can be used. Transmission capacity log ₂ (1 + SNR) obtained from SNR
Can be used.

  For example, it is conceivable to perform control so that each evaluation value is maximized using a certain evaluation value Q that can be expressed as follows.

  SNR 'is the expected communication quality.

FIG. 5 shows an example in which a communication sequence and an interference space are selected using the square value of the inner product value. The square value of the inner product value corresponds to a power value that decreases when the orthogonal spatial channel response matrix is created, and ρ ² can be used as an evaluation value indicating that the power value of the channel is decreased.

The eigenvalue-to-noise ratio of each communication channel is shown in dB in the upper part of the table of FIG. This represents the maximum SNR that can be expected when there is no interference power, but an interference spatial channel response matrix is defined to prevent interference with other channels, and ρ ² shown below is used when all streams are used. And the estimated communication quality [dB] significantly deteriorates. Therefore, it is possible to select the communication sequence and interference space to achieve the maximum spectral efficiency from the sum of the [rho ^2.

6 is an example of reducing the communication path of each communication partner so that the sum of [rho ² as a selection method is 0.5 or less. By selecting in this way, each communication series can be used effectively.

FIG. 7 shows the estimated signal-to-noise ratio Γv _{, l} at the time of multiuser transmission of the l-th communication sequence of the v-th communication partner, from the estimated signal quality SNR ′ and the inner product value described in the upper row of the estimated quality table. It is also possible to select an interference space basis vector candidate that maximizes C expressed by the following expression.

Alternatively _, it is possible to select the modulation scheme to be applied from the estimated signal-to-noise ratio Γv _{, l} and then select the bit error to be the smallest.

For example,
Γv _{, l} = (1-Σζρ ² ) · SNRv _{, l}
As described above, an estimated signal-to-noise ratio at the time of multi-user transmission can be used.

  FIG. 7 shows an example when priority is given to the transmission capacity to the communication partner B. In this way, communication sequences of other communication partners can be reduced so that the estimated transmission capacity to any communication partner is increased.

In addition, when the candidate to be selected is determined from the interference space basis vector candidates as described above, the interference space channel vector is used for the kth communication partner using the interference space basis vector candidates other than the kth communication partner. Response matrix G _(k)
G _(k) = [Ω ' ₁₎ ^T , , Ω ′ _(k−1) ^T _, Ω ′ _{(k + 1)} ^T ,. ' _(Mu) ^T ] ^T
To decide.

Using the interference space channel response matrix G _(k) , the interference space basis vector for the k-th communication partner can be determined in the same manner as Equation (8). Here, Ω ′ _(l) can define the reception weight formed by the l-th communication partner as the initial weight when the interference space basis vector candidate selected for the l-th communication partner is used as the transmission weight. ,
Ω ' _(l) = F _(l) ^H H _(l)
(L ′ _(l) × M _T matrix). L ′ _(l) is the number of interference space basis vectors selected for the l-th communication partner, and is equal to or greater than the number L _(l) of the l-th communication sequence.

When no candidate is selected, the number of interference space basis vectors L ′ _(l) is L ′ _(l) = 0, and the interference space basis vector candidate Ω ′ _(l) is not used.

At this time the interference spatial channel response matrix _{G (k) is'} has a _(k) × _{M T} matrix, L 'L'_'(k) from the number L' _all of the selected interference spatial basis vectors candidate The number is _obtained by subtracting L ′ _(k) . From this interference space channel response matrix, interference space basis vector group E _(k)
E _(k) = [e _{k, 1} ^T , e _{k, 2} ^T ,..., E _{k, L2 (k)} ^T ] ^T
And the transmission weight can be determined from the orthogonal channel response matrix.

  Alternatively, several candidates of the interference space basis vector candidate selection method are output to the orthogonal space arithmetic circuit 206, the orthogonal spatial channel response matrix is obtained for each candidate, and the actual result is calculated from the calculation result of the transmission weight arithmetic circuit 205. Among the characteristics obtained when the transmission weight is used, the one having the highest communication speed can be selected. Alternatively, priority can be given to increasing the communication speed to one or more specific communication partners.

In addition, when performing output from the interference space candidate calculation circuit 209 to the communication space selection circuit 207, the interference space basis vector candidates that have not been selected by each communication partner are input to the communication space selection circuit 207 as quasi-interference space basis vector candidates. Can be output.

Let the number of quasi-interference space basis vector candidates selected for the kth communication partner be L ⁺ _(k) . Here, the number of quasi-interference space basis vector candidates defined for communication partners other than the kth communication partner is L ⁺ ′ _(k) . When attention is paid to the kth communication partner, the L ⁺ ′ _(k) quasi-interference space basis vector candidates output in this way are interference spatial channel response matrices G _(l) (L ₎ of communication partners other than the kth. '' _(L) × M _T matrix) and an extended interference spatial channel response matrix G ′ _{(l)
^{G '(l) = [G}} (l) T, G + (l) T] T ((L''(k) + L +' (k)) × M T matrix)
Can be defined.

Here, when representing the quasi-interference spatial basis vectors candidate selected as ^{_{Ω '' (k) (L}} + (k) × M T matrix),
G ⁺ _(l) = [Ω ″ ₍₁₎ ,..., Ω ″ _(k−1) , Ω ″ _{(k + 1)} ,..., Ω ″ _(Mu) ] (L ⁺ ′ _{( k)} × M _T matrix). Using this extended interference spatial channel response matrix G ′ _(l) , the interference space basis vector group E _(k)
E _(k) = [e _{k, 1} ^T , e _{k, 2} ^T ,..., E _{k, L} + _{(k) + L2 (k)} ^T ] ^T
To decide.

  Using this interference spatial channel response matrix, the orthogonal spatial channel response matrix can be expressed as:

Here, ρ _i (i ≦ L ″ _(k) ) corresponding to the interference space basis vector is 1, and ρ _i (L ″ _(k) + 1 ≦ i ≦ L ⁺ _{(k )} corresponding to the quasi-interference basis. ₎ + L ″ _(k) ) is set to a value larger than 0 and smaller than 1, so that the orthogonality condition is not given completely and the amount of interference can be reduced.

  FIG. 8 shows a transmission flow in the second embodiment of the present invention. In FIG. 8, before performing communication, a channel response matrix corresponding to a communication partner performing transmission is estimated in advance (step S201).

When the communication partner and transmission data are determined (step S202), using the channel response matrix of the corresponding communication partner, interference space basis vector candidates are calculated for each communication partner (step S203), and correspond to different communication partners. An interference space basis vector group, or an interference space basis vector group and a quasi-interference space basis vector group are determined using the inner product value of the vector candidates to be processed (step S204). An orthogonal spatial channel response matrix is calculated using the interference spatial basis vector and the channel response matrix (step S205). A transmission weight and a transmission mode to be applied are determined using the orthogonal spatial channel response matrix determined for each communication partner (step S206). The transmission data is input (step S207), the data to be transmitted is divided into L series with the bit length corresponding to the modulation method (step S208), modulation processing and addition of a known signal are performed (step S209), and signal conversion is performed. The processing in the circuit is performed on a symbol-by-symbol basis. For example, if the signal vector at the symbol for the kth communication partner is S _k and the transmission weight for the _kth communication partner is W _k , S _k → W _k · S _k The signal conversion process for performing the signal conversion process is performed (step S210). In this way, the processed baseband signal is converted into an RF signal by a radio unit corresponding to each antenna and transmitted (step S211).

(Third embodiment)
FIG. 9 shows a configuration example of the transmission weight determination block 110 in the third embodiment of the present invention. In FIG. 9, 305 is a transmission weight calculation circuit, 306 is an orthogonal space calculation circuit, 307 is a communication space selection circuit, 308 is a channel response matrix acquisition circuit, 309 is an interference space candidate calculation circuit, and 310 is an additional interference space selection circuit. .

In the third embodiment of the present invention, the additional interference space selection circuit 310 considers a vector space other than the initial weight considered for the communication partner and channel time variation, and transmits the transmission weight to the other communication partner as an interference signal. , Γ _{1 ″ add (k)} or additional quasi-interference space basis vectors γ ′ ₁ ,..., Γ ′ _L + _{add (k)} Is output to the communication space selection circuit 307. Additional interference spatial basis vectors number L _{'' add (k),} the additional quasi interference spatial basis vectors number is ^L _{+ add (k).}

The extended interference spatial channel response matrix G ′ _(k) for the k th communication partner is
(L ″ _(k) + L ″ _{add (k)} + L ⁺ _(k) + L ⁺ _{add (k)} ) × M _T
It becomes a matrix and can be expressed as follows.
G ′ _(k) = [G _(l) ^T _, γ ₁ ^T ,..., Γ _{L ″ add (k)} ^T _, G ⁺ _(l) ^T _, γ ′ ₁ ^T _,. + _{Add (k)} ^T ] ^T

^{_{γ 1 T, ···, γ L2add}} (k) T is the additive interference space basis vectors ^{_{group, γ '1 T, ···,}} γ' L + add (k) T is added quasi interference space basis vectors group It is.

Using this extended interference spatial channel response row G _'(l), as in Equation (8), the interference spatial basis vectors group _{E (k)}
E _(k) ⁼ determine ^{_{[e k, 1 T, e}} k, 2 T, ··· e k, L2 (k) + L2add (k) + L + (k) + L + add (k) T] T To do.
Using this interference spatial channel response matrix, the orthogonal spatial channel response matrix can be expressed as:

Here, ρ _i (i ≦ (L ″ _(k) + L ″ _{add (k)} )) corresponding to the interference space basis vector is set to 1, and ρ _i (L ″ (L ″ _{( k)} + L ″ _{add (k)} + 1 ≦ i ≦ L ″ _(k) + L ″ _{add (k)} + L ⁺ _(k) + L ⁺ _{add (k)} ) is set to a value greater than 0 and less than 1 By doing so, the orthogonality condition is not given completely, and the amount of interference can be reduced.

When the transmission weight is determined using the information of the orthogonal spatial channel response matrix H ′ _(k) determined in this way, a transmission beam that is robust against time variation can be formed. As the additional interference space, changes due to time fluctuations are estimated from the channel response matrix of the near frequency band in frequency orthogonal wave division multiplexing (OFDM) or the past channel response matrix obtained until transmission is performed, and at the time of transmission, An expected channel response matrix can be used.

  Alternatively, when a communication partner (interfered communication partner) who does not want to give an interference signal other than the communication partner of the transmission / reception apparatus is within a range where the transmission power can reach, the channel response matrix for the interfered communication partner is added to the additional interference space. Alternatively, it can be defined as an additional quasi-interference space.

  FIG. 10 shows a transmission flow in the third embodiment of the present invention. In FIG. 13, before performing communication, a channel response matrix corresponding to a communication partner performing transmission is estimated in advance (step S301).

When the communication partner and transmission data are determined (step S302), an interference space basis vector candidate is calculated for each communication partner using the channel response matrix of the corresponding communication partner (step S303). An inner space value of vector candidates corresponding to different communication partners is used to determine an interference space basis vector group, or an interference space basis vector group and a quasi-interference space basis vector group (step S304). The change is estimated, an additional interference space basis vector or an additional quasi-interference space basis vector is determined, and added as these additional interference space basis vectors (step S305). An orthogonal spatial channel response matrix is calculated using the interference space basis vector, the quasi-interference space basis vector, the additional interference space basis vector, the additional quasi-interference space basis vector, and the channel response matrix (step S306). Using the orthogonal spatial channel response matrix determined for each communication partner, the transmission weight and transmission mode are determined (step S307), the transmission data is input (step S308), and the data to be transmitted is converted to the bit corresponding to the modulation method. The length is divided into L series (step S309), modulation processing and addition of a known signal are performed (step S310), and the processing in the signal conversion circuit is performed on a symbol-by-symbol basis. If the signal vector is S _k and the transmission weight for the k th communication partner is W _k , a signal conversion process for performing a signal conversion process of S _k → W _k · S _k is performed (step S311). It is converted into a radio signal and transmitted from the antenna (step S312).

(Fourth embodiment)
FIG. 11 shows a configuration example of the transmission weight determination block 110 in the fourth embodiment of the present invention. In FIG. 11, reference numerals 405-1 and 405-2 denote transmission weight calculation circuits, 406 an orthogonal space calculation circuit, 407 a communication space selection circuit, 409 an interference space candidate calculation circuit, 410 an additional interference space selection circuit, and 411 a transmission. The rank setting circuit is shown.

  In the fourth embodiment of the present invention, when a communication partner and transmission data are determined, the transmission rank setting circuit 411 assumes a first communication partner (μth communication partner) that preferentially constructs a communication sequence. ) And the first communication sequence number (assuming Λ) are output to the channel information acquisition circuit 408.

The channel information acquisition circuit 408 outputs the channel response matrix information of the designated first communication partner to the first transmission weight calculation circuit 405-1, and the first transmission weight calculation circuit 405-1 is input. The transmission weights w ^f _{(μ, 1)} to w ^f _{(μ, Λ)} of the first communication sequence number are determined by linear calculation from the channel response matrix of the first communication partner, and an interference space candidate calculation circuit Output to 409.

As a method of determining the first transmission weight, for example, the transmission side eigenvector V ₍ μ ₎ obtained by singular value decomposition of the channel response matrix H ₍ μ ₎ is used in the order of corresponding eigenvalues in order of L eigenvectors, The channel response matrix transposed matrix H ₍ μ ₎ ^H uses a part of the unitary matrix determined by the orthogonalization method as the column vector, or obtains a vector highly correlated with the eigenvector by some means. The transmission weights can be used as w ^f _{(μ, 1)} to w ^f _{(μ, P)} .

The interference space candidate calculation circuit 409 uses the received weight formed by the μ-th communication partner when the input first transmission weights w ^f _{(μ, P)} to w ^f are used as transmission weights as the initial weight F _{f (} the interference spatial channel response matrix as _mu), added ^{_{F f (μ) H H (}} μ), the third same calculation path and following embodiments, a second transmission weight other than mu th communication partner, The second transmission weight of the μ-th communication partner is calculated.

At this time, the second transmission weight other than the μ-th communication partner can be determined by the same flow as that of the third embodiment, but the μ-th communication partner is set in the interference spatial channel response matrix. It is necessary to add a channel response vector between the reception beam and the transmission antenna by the reception weight for the first transmission weights w ^f _{(μ, 1) to} w ^f _{(μ, W)} determined as the communication partner. By controlling in this way, the second transmission weight determined as the μ-th communication partner can be orthogonal to the first transmission weight of the μ-th communication partner.

  The first transmission weight obtained in this way can obtain a high transmission capacity corresponding to the eigenvalue obtained by the original channel response matrix. In addition, since the communication sequence based on the first transmission weight becomes an interference wave other than the μ-th communication partner, these communication partners need to remove this interference signal at their terminals.

  FIG. 12 shows a transmission flow in the fourth embodiment of the present invention. In FIG. 12, before performing communication, a channel response matrix corresponding to a communication partner performing transmission is estimated in advance (step S401).

When the communication partner and transmission data are determined (step S402), first, the first communication partner is selected from among them (step S403). The first transmission weight and modulation scheme for the first communication partner are determined from the channel response matrix of the first communication partner (step S404). In order to determine a communication sequence other than the first communication sequence of the first communication partner and a transmission weight for the other communication partner, an interference space basis vector candidate is calculated for each communication partner from the channel response matrix (step S405). At this time, the first transmission weight is always included in the interference space basis vector candidate of the first communication partner. An inner space value of vector candidates corresponding to different communication partners is used to determine an interference space basis vector group, or an interference space basis vector group and a quasi-interference space basis vector group (step S406). To increase stability, estimate the channel response at the time of transmission from the information of the past channel response matrix, determine the additional interference space basis vector or additional quasi-interference space basis vector obtained from these channel responses, and This is added as an additional interference space basis vector (step S407). An orthogonal spatial channel response matrix is calculated using the interference space basis vector, the quasi-interference space basis vector, the additional interference space basis vector, the additional quasi-interference space basis vector, and the channel response matrix (step S408). Using the orthogonal spatial channel response matrix determined for each communication partner, the transmission weight and the transmission mode are determined (step S409), the transmission data is input (step S410), and the data to be transmitted is changed to the bit corresponding to the modulation method. The length is divided into L series (step S411), modulation processing and addition of a known signal are performed (step S412), and the processing in the signal conversion circuit is performed on a symbol-by-symbol basis. If the signal vector of S _k is S _k and the transmission weight for the k th communication partner is W _k , a signal conversion process for performing a signal conversion process of S _k → W _k · S _k is performed (step S413), and the radio unit Is converted into a radio signal and transmitted from the antenna (step S414).

(Fifth embodiment)
FIG. 13 shows a configuration example of the transmission weight determination block 110 in the fifth embodiment of the present invention. In the figure, 505-1 is a first transmission weight arithmetic circuit, 505-2 is a second transmission weight arithmetic circuit, 506-1 is a first orthogonal space arithmetic circuit, and 506-2 is a second orthogonal space arithmetic circuit. , 507-1 is a first communication space selection circuit, 507-2 is a second communication space selection circuit, 509-1 is a first interference space candidate calculation circuit, and 509-2 is a second interference space candidate calculation circuit. , 510 indicates an additional interference space selection circuit, and 511 indicates a transmission rank setting circuit.

In the fifth embodiment of the present invention, when the first transmission weight is used for a plurality of communication partners, it is necessary to define an orthogonal spatial channel response matrix for the first transmission weight and determine the transmission weight. There is. It is assumed that the communication partners performing priority transmission are μ-th and η-th, and the first communication sequence numbers for the μ-th and η-th communication partners are Λ ( _μ ) and Λ ( _η ).

In the transmission rank setting circuit 511, communication partners (μth and ηth communication partners) that preferentially improve communication quality and the number of communication sequences Λ ( _μ ) and Λ ( _η ) of the first communication series are determined. To the channel information acquisition circuit 508.

The first interference space candidate computation circuit 509-1 uses the channel response matrixes of the η th and μ th communication partners, respectively, to obtain a first interference space basis vector candidate Ω ^f _(η) = [ω ^f _{η, 1} ^T , ω ^f _{η, 2} ^T , ... ω ^f _{η, MR (η)} ^T ] ^T ,
Ω ^f _(μ) = [ω ^f _{μ, 1} ^T , ω ^f _{μ, 2} ^T , ... ω ^f _{μ, MR (μ)} ^T ] ^T
Calculate.

Interference spatial basis vectors candidate channel response matrix H _(eta), obtained in the linear operation from H _(mu), a vector. For example, the channel response matrix H _(k) is subjected to singular value decomposition, and H _(k) = U _(k) D _(k) V _(k) ^H
Eigenvector V _(k)
V _(k) = (vk _{, 1} ,..., Vk _{, MR (k)} )
Can be selected as ω _{k, j} .

The inner product value is evaluated between the row vectors of the interference space basis vector candidates Ω ^f _(η) and Ω ^f _(μ) obtained here, and similarly to the interference space candidate calculation circuits of the second to fourth embodiments, An interference space basis vector and a quasi-interference space basis vector are selected, and a first interference space channel response matrix G ^f _(η)
G ^f _(η) = [Ω ^f ′ _(μ) ^T ] ^T is determined.

Here, when the interference space basis vector candidate selected for the μ th communication partner is used as the transmission weight, Ω ^f ′ _(μ) uses the reception weight formed by the μ th communication partner as the initial weight F ^f _(μ). Used as
Ω ^f _'(μ) = a _{^{F f (μ) H H (}} μ) and then can be obtained matrix, ^{L f'} is a _(μ) × _{M T} matrix. L ^f ′ _(μ) is the number of interference space basis vectors selected for the μ th communication partner.

Using the first interference space channel response matrix G ^f ( _η ), the first interference space basis vector E ^f _(η) for the η th communication partner as in Equation 8.
E ^f _(η) = [e ^f _{η, 1} ^T , e ^f _{η, 2} ^T , ... e ^f _{η, L2 (η)} ^T ] ^T
Can be determined.

At this time the interference spatial channel response matrix ^{G f} _{(eta) ^{is, L f '' (η)}} × has a _{M T} matrix, ^{L f} '' _^(η) is the number of first interference spatial basis vectors candidate The number is obtained by subtracting L ′ _(η) from L ′ _all .

Using the first interference spatial basis vector group, the channel response matrix H _(eta) from the first orthogonal spatial channel response matrix H ^f 'a _(eta) can be determined by Equation (9), the resulting A vector obtained by linear calculation from one orthogonal spatial channel response matrix H ^f ′ _(η) is defined as a first transmission weight.

For example, singular value decomposition of the channel response matrix H ^f ′ _(η)
H ^f ′ _(η) = U ^f ′ _(η) D ^f ′ _(η) V ^f ′ _(η) ^H
Eigenvector V ^f ′ _(η)
V ^f ′ _(η) = (v ^f ′ _{η, 1} ,..., V ^f ′ _{η, MR (η)} )
Can be the first transmission weight. Similarly, the first transmission weight for the μ-th communication partner can be obtained.

  At this time, in the first interference space candidate calculation circuit 509-1, the channel information acquisition circuit 508 outputs a channel response matrix of a communication partner other than the first communication partner to the first interference space candidate calculation circuit. You can also. The interference space basis candidate vector candidate is also calculated from the channel response matrix of the second communication partner input at this time.

  Then, when selecting the first interference space basis vector group for the first communication partner, the inner product value with the interference space basis vector candidate formed for the second communication partner can also be considered.

  Having a high inner product value with the interference space basis vector candidate formed in the second communication partner means that it becomes difficult to remove the signal transmitted to the first communication partner in the communication partner. Therefore, by not selecting such an interference space basis vector candidate, transmission to this space can be prevented.

It is also possible to determine by combining a plurality of first interference space basis vector candidates. It is assumed that three first interference space basis vector candidates are calculated from H ( _μ ) for the μ-th communication partner, and two of these are selected. Here, when the second interference space basis vector ω _{μ, 2} has an absolute value of a high inner product value with respect to the second interference space basis vector group, the second interference space basis vector ω _{μ, 2} is the other communication partner. The amount of interference is expected to be large.

However, if it is considered that the space formed by the second interference space basis vector ω _{μ, 2} can form a communication sequence having a communication quality higher than that of the third interference space basis vector ω _{μ, 2, the} space for the μ th communication partner The second interference space basis vector candidate can be used as follows.

Here, χ is 0 <χ <1.

Also, the first interference space basis vector candidate not selected as the first interference space basis vector is selected as the first quasi-interference space basis vector, and the first extended interference space channel response matrix G ^f ′ _{(η )} As follows,
G ^f ′ _(η) = [G ^f ′ _(η) , G ^{f +} _(η) ]
As described in the second embodiment, the inner product value can also be reduced although it is not completely orthogonal.

When G ^{f +} _(eta) is expressed as a quasi-interference spatial basis vectors candidate selected ^{_{Ω '' (μ) (L}} f + (μ) × M T matrix),
_{^{G + (η) = [Ω}} f '' (μ)] (L f + '(k) × M T matrix)
It can be expressed as.

  FIG. 14 shows a transmission flow in the fifth embodiment of the present invention. Before performing communication, a channel response matrix corresponding to a communication partner performing transmission is estimated in advance (step S501), and an additional interference space is selected (step S502).

When a communication partner and transmission data are determined (step S503), a plurality of first communication partners are first selected from them (step S504). In the first communication partner, the first interference space basis vector candidate, or the first interference space basis vector and the second interference space basis vector candidate are calculated (step S505), and correspond to different first communication partners. Using an inner product value of vector candidates, or using an inner product value of vector candidates corresponding to different first communication partners and a second interference space basis vector candidate corresponding to the second communication partner, A first interference space basis vector group, a quasi-interference space basis vector, or both are determined (step S506). From the channel response matrix of the first communication partner, the interference space basis vector, and the quasi-interference space basis vector, One orthogonal spatial channel response matrix is calculated (step S507). A first transmission weight is calculated from the obtained first orthogonal spatial channel response matrix to determine a corresponding transmission mode (step S508). Next, in order to determine a communication sequence other than the first communication sequence of the first communication partner and a transmission weight for the second communication partner, an interference space basis vector candidate for each communication partner is determined from the corresponding channel response matrix. Is calculated (step S509). At this time, the first transmission weight is always included in the interference space basis vector candidate of the first communication partner. A second interference space basis vector group, or a second interference space basis vector group and a second quasi-interference space basis vector group are determined using inner product values of vector candidates corresponding to different communication partners (step S510). In addition, when improving the stability of communication quality against time variation, the channel response at the time of transmission is estimated from the information of the past channel response matrix, and the second additional interference space basis vector obtained from these channel responses or A second additive quasi-interference space basis vector is determined. Second orthogonal spatial channel response using second interference space basis vector, second quasi-interference space basis vector, second additional interference space basis vector, second additional quasi-interference space basis vector, and channel response matrix A matrix is calculated (step S511). The second transmission weight and the transmission mode to be applied are determined using the second orthogonal spatial channel response matrix determined for each communication partner (step S512), the transmission data is input, and the data to be transmitted is set to the transmission mode. The signal is divided into L series with the corresponding bit length (step S513), modulation processing and addition of a known signal are performed (step S514), and the processing in the signal conversion circuit is performed in symbol units, for example, the k-th communication the signal vector at symbol to the counter and S _k, when a transmission weight for the k th communication partner and W _k, and performs signal conversion processing to carry out signal conversion processing of _{S k → W k · S k} ( step S515 ), Converted into a radio signal by the radio unit, and transmitted from the antenna (step S516).

(Sixth embodiment)
FIG. 15 shows a configuration example of the transmission weight determination block 110 in the sixth embodiment of the present invention. In FIG. 15, 605-1 is a first transmission weight calculation circuit, 605-2 is a second transmission weight calculation circuit, 606-1 is a first orthogonal space calculation circuit, and 606-2 is a second orthogonal space calculation. Circuit, 607-1 is a first communication space selection circuit, 607-2 is a second communication space selection circuit, 609-1 is a first interference space candidate calculation circuit, and 609-2 is a second interference space candidate calculation. 610-1 is a first additional interference space selection circuit, 610-2 is a second additional interference space selection circuit, and 611 is a transmission rank setting circuit.

In the transmission rank setting circuit 612, communication partners (μth and hth communication partners) that preferentially improve communication quality and the number of communication sequences L ( _m ) and L ( _h ) of the first communication series are determined. And output to the channel information acquisition circuit 608.

  When the first transmission weight is used for one or a plurality of communication partners, an orthogonal spatial channel response matrix is defined for the first transmission weight, and the transmission weight is determined. Here, as an example, the communication partners that perform priority transmission are the m-th and the h-th.

  In the transmission rank setting circuit 612, communication partners (μth and ηth communication partners) whose communication quality is improved with priority are determined and output to the first interference space candidate calculation circuit 609-1.

The first interference space candidate computation circuit 609-1 uses the channel response matrixes of the η th and μ th communication partners, respectively, to obtain a first interference space basis vector candidate Ω ^f _(η) = [ω ^f _{η, 1} ^T , ω ^f _{η, 2} ^T , ... ω ^f _{η, MR (η)} ^T ] ^T ,
Ω ^f _(μ) = [ω ^f _{μ, 1} ^T , ω ^f _{μ, 2} ^T , ... ω ^f _{μ, MR (μ)} ^T ] ^T
Is calculated.

Interference spatial basis vectors candidate channel response matrix H _(eta), obtained in the linear operation from H _(mu), a vector. Or singular value decomposition of the channel response matrix H _(k) ,
H _(k) = U _(k) D _(k) V _(k) ^H
Eigenvector V _(k)
V _(k) = (vk _{, 1} , ..., vk _{, MR (k)} )
Can be selected as ω _{k, j} .

The inner product value is evaluated between the row vectors of the interference space basis vector candidates Ω ^f _(η) and Ω ^f _(μ) obtained here, and similarly to the interference space candidate calculation circuits of the second to fourth embodiments, Interference space basis vectors and quasi-interference space basis vectors can be formed.

  At this time, the first interference space candidate calculation circuit 609-1 can also output the channel response matrix of a communication partner other than the first communication partner to the first interference space candidate calculation circuit 609-1. The interference space basis candidate vector candidate is also calculated from the channel response matrix of the second communication partner input at this time.

  Then, when selecting the first interference space basis vector group for the first communication partner, the inner product value with the interference space basis vector candidate formed for the second communication partner can also be considered. Having a high inner product value with the interference space basis vector candidate formed in the second communication partner means that it becomes difficult to remove the signal transmitted to the first communication partner in the communication partner. Therefore, by not selecting such an interference space basis vector candidate, transmission to this space can be prevented.

At this time, if there is a communication partner (n-th communication partner) that cannot remove the signal of the first communication series because the interference is too high, the n-th communication partner is set as a communication partner that cannot perform interference. , The corresponding channel response matrix H ( _n ), or at least a part thereof, is output to the first additional interference space selection circuit 610-1.

In the first additional interference space selection circuit 607-1, interference space basis vector candidates ω _{n, 1} , ω _{n, 2} ,... Ω _n, obtained by linear calculation from the channel response matrix for the n-th communication partner _{. MR (n)} or the reception weight formed by the nth communication partner when any one of them is used as the transmission weight is used as the initial weight F _{f (n)} , and Ω ^f ′ (n) = F ^f _(n) ^H H _(n) is output to the first communication space selection circuit 606-1 in addition to the first additional interference space basis vector or the first additional quasi-interference space basis vector. The first additional interference space basis vector number is L ^f ″ _{add (k)} , and the additional quasi-interference space basis vector number is L ^{f +} _{add (k)} .

First extended interference spatial channel response matrix ^{G f} _'(k) is, ^{(L f' _{becomes' (k) + L f '}} ' add (k) + L f + (k) + L f + add (k)) × MT matrix,
_{^{G f '(k) = [}} G f (l) T, g f 1 T, ···, g f L2add (k) T, G f + (l) T, g f' 1 T, ···, g _{^{f 'L + add (k)}} T] T
It can be expressed as. Using this first extended interference spatial channel response row G ^f ′ _(l) , the interference space basis vector group E ^f _(k)
E ^f _(k) = e ^f _{k, 1} ^T , e ^f _{k, 2} ^T ,..., E ^f _k , _L f _{2 (k) + L} f _{2add (k) + L} f _{+ (k) + L} f _{+ add (k )} ^T ] Determine ^T.

  Using this interference spatial channel response matrix, the first orthogonal spatial channel response matrix can be expressed as:

Here, ρ _i (i ≦ (L ^f ″ _(k) + L ^f ″ _{add (k)} )) corresponding to the interference space basis vector is set to 1, and ρ _i (L ^f corresponding to the quasi-interference basis is set. ^{_{'' (k) + L f}} '' add (k) + 1 ≦ i ≦ L f '' (k) + L f '' add (k) + L f + (k) + L f + add (k)) is 0 By setting the value larger and smaller than 1, it is possible to reduce the amount of interference without giving the orthogonal condition completely.

When the transmission weight is determined using the information of the orthogonal spatial channel response matrix H ^f ′ _(k) determined in this way, it is possible to avoid interference with a communication partner whose interference with the first communication sequence is difficult to remove. . Alternatively, as an additional interference space, a change due to time variation is estimated from a channel response matrix of a near frequency band in frequency orthogonal wave division multiplexing (OFDM) or a past channel response matrix obtained until transmission is performed. By using the channel response matrix expected in, directivity control that is robust against time variation can be performed.

  Alternatively, the communication partner whose interference is difficult to remove is at least one of the number of antenna elements owned by the communication partner other than the first communication partner, the decoding algorithm, the consumable power, and the average received power in the channel information acquisition circuit. One or a plurality of non-interference communication partners can be determined.

  FIG. 16 shows a transmission flow in the sixth embodiment of the present invention. In FIG. 6, before performing communication, a channel response matrix corresponding to a communication partner to be transmitted is estimated in advance (step S601), a first additional interference space is selected (step S602), and a second additional interference space is selected. Is selected (step S603).

  When the communication partner and transmission data are determined (step S604), first, a plurality of first communication partners are selected from them (step S605). In the first communication partner, a first interference space basis vector candidate is calculated (step S606), and using the inner product value of the vector candidates corresponding to different first communication partners, the first interference space basis vector group, A quasi-interference space basis vector or both are determined (step S607), and a first orthogonal spatial channel response matrix is calculated from the channel response matrix of the first communication partner, the interference space basis vector, and the quasi-interference space basis vector. (Step S608). A first transmission weight is calculated from the obtained first orthogonal spatial channel response matrix to determine a corresponding transmission mode (step S609).

Next, in order to determine a communication sequence other than the first communication sequence of the first communication partner and a transmission weight for the second communication partner, an interference space basis vector candidate for each communication partner is determined from the corresponding channel response matrix. Is calculated (step S610). At this time, the first transmission weight is always included in the interference space basis vector candidate of the first communication partner. An inner space value of vector candidates corresponding to different communication partners is used to determine an interference space basis vector group, or an interference space basis vector group and a quasi-interference space basis vector group (step S611), and communication quality against time variation is determined. To increase stability, estimate the channel response at the time of transmission from the information of the past channel response matrix, determine the additional interference space basis vector or additional quasi-interference space basis vector obtained from these channel responses, and additional interference spatial added as basis vectors.
An orthogonal spatial channel response matrix is calculated using the interference space basis vector, the quasi-interference space basis vector, the additional interference space basis vector, the additional quasi-interference space basis vector, and the channel response matrix (step S612). Using the orthogonal spatial channel response matrix determined for each communication partner, the transmission weight and the transmission mode to be applied are determined (step S613), the transmission data is input (step S614), and the data to be transmitted corresponds to the transmission mode. The data is divided into L series by the bit length (step S615), modulation processing and addition of a known signal are performed (step S616), and the processing in the signal conversion circuit is performed in symbol units, for example, a symbol for the kth communication partner. the signal vector at the _{S k,} when a transmission weight for the k th communication partner and _{W k,} and performs signal conversion processing to carry out signal conversion processing of _{S k → W k · S k} ( step S617), the radio Is converted into a radio signal by the unit and transmitted from the antenna (step S618).

In any case of the first to sixth embodiments of the present invention, after the transmission weight is determined by the transmission weight calculation circuit, the transmission weight information is output to the interference space calculation circuit or the interference space candidate calculation circuit. Then, these transmission weights are selected as interference space basis vector candidates, interference space basis vectors and quasi-interference space basis vectors are determined again, and the process of determining the transmission weight is repeated any number of times, and the resulting transmission weight is finally determined. It can also be used as a typical transmission weight. It is also possible to determine the transmission weight so that no interference occurs in the interference space basis vector without obtaining the orthogonal spatial channel response matrix.

  The above repetitive calculation is the same in the first transmission weight calculation circuit 605-1 and the second transmission weight calculation circuit 605-2, and is output to the corresponding interference space calculation circuit or interference space candidate calculation circuit. The transmission weight can be determined again by the same processing.

  FIG. 17 shows the frequency utilization efficiency when using the embodiment of the present invention. When the number of transmitting elements is 8, the number of communication partners is 2 terminals, the number of receiving elements is 4 elements, and the total power is transmitted by the omnidirectional antenna in the transmitting apparatus, the average received SNR at each receiving antenna of the communicating station Is 30 dB. At this time, (1) when performing 8 × 4 MIMO communication at different times for each communication partner, (2) using four communication sequences for the first communication partner, When performing multi-user MIMO communication using a communication sequence, (3) using four communication sequences for the first communication partner and calculating the sum of the square values of the inner product values for the second communication partner. The eigenvector having the largest sum other than the first eigenvector is selected not to be used. Where i. i. d. The channel model is used, the reception tap interval is 50 nsec, the reception signal of 16 taps is used, Rayleigh fading is assumed for each reception signal, the expected value of the power distribution is given as an index, and the delay spread is 100 nsec. Set. The frequency utilization efficiency achievable by the three methods is shown in FIG. As shown in FIG. 13, it can be confirmed that the highest transmission capacity is achieved in the case of (3).

  In the fourth to sixth embodiments of the present invention, a reception beam and a transmission antenna formed by the first communication partner for the first transmission weight are used as the interference spatial channel response matrix for determining the second transmission weight. In addition to a channel response vector between and a reception beam formed by the second communication partner for the first transmission weight, a channel response vector between and can be included. The second communication partner to be included in the interference spatial channel response matrix can be arbitrarily set, and by adding this condition to the interference space, the first communication generated in the second communication partner in the transmission using the second transmission weight. Interference from the first transmission weight of the partner can be reduced.

  INDUSTRIAL APPLICABILITY The present invention can be used to perform MIMO communication by using one or a plurality of signal sequences addressed to a plurality of communication partner stations on the same frequency channel using spatial multiplexing at the same time.

It is a figure which shows the structural example of the transmission part of the 1st Embodiment of this invention. It is a figure which shows the transmission flow in the 1st Embodiment of this invention. It is a figure which shows the structural example of the transmission weight determination block in the 2nd Embodiment of this invention. It is a figure which shows the example of selection of the communication sequence number and interference space by the square value of an inner product value. It is a figure which shows the example of selection of the communication sequence number and interference space by the square value of an inner product value. It is a figure which shows the example of selection of the communication sequence number and interference space by the square value of an inner product value. It is a figure which shows the example of selection of the communication sequence number and interference space by the square value of an inner product value. It is a figure which shows the transmission flow in the 2nd Embodiment of this invention. It is a figure which shows the structural example of the transmission weight determination block in the 3rd Embodiment of this invention. It is a figure which shows the transmission flow in the 3rd Embodiment of this invention. It is a figure which shows the structural example of the transmission weight determination block in the 4th Embodiment of this invention. It is a figure which shows the transmission flow in the 4th Embodiment of this invention. It is a figure which shows the structural example of the transmission weight determination block in the 5th Embodiment of this invention. It is a figure which shows the transmission flow in the 5th Embodiment of this invention. It is a figure which shows the structural example of the transmission weight determination block in the 6th Embodiment of this invention. It is a figure which shows the transmission flow in the 6th Embodiment of this invention. It is a figure which shows the effect of embodiment of this invention. It is a figure which shows the structural example of the radio station in a prior art.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Data division circuit 101-1 to 101-L ... Modulation circuit 102 ... Transmission signal conversion circuit 103-1 to 103-3 ... Radio | wireless part 104-1 to 104-MT ... Antenna 105: Transmission weight calculation circuit 106 ... Orthogonal space calculation circuit 107 ... Interference space calculation circuit 205 ... Transmission weight calculation circuit 206 ... Orthogonal space calculation circuit 207 ... Interference space calculation circuit 208 Channel information acquisition circuit 305 Transmission weight calculation circuit 306 Orthogonal space calculation circuit 307 Interference space calculation circuit 308 Channel information acquisition circuit 309 Interference space candidate calculation circuit 310 Additional interference space selection circuit 405-1: first transmission weight calculation circuit 405-2: second transmission weight calculation circuit 406: orthogonal space calculation times 407 ... Communication space selection circuit 408 ... Channel information acquisition circuit 409 ... Interference space candidate calculation circuit 410 ... Additional interference space selection circuit 411 ... Transmission rank setting circuit 505-1 ... First Transmission weight arithmetic circuit 505-2 ... second transmission weight arithmetic circuit 506-1 ... first orthogonal space arithmetic circuit 506-2 ... second orthogonal space arithmetic circuit 507-1 ... 1st communication space selection circuit 507-2 ... 2nd communication space selection circuit 508 ... Channel information acquisition circuit 509-1 ... 1st interference space candidate calculation circuit 509-2 ... 2nd Interference space candidate calculation circuit 510... Additional interference space selection circuit 511... Transmission rank setting circuit 605-1... First transmission weight calculation circuit 605-2. -1. First orthogonal space arithmetic circuit 606-2 ... second orthogonal space arithmetic circuit 607-1 ... first communication space selection circuit 607-2 ... second communication space selection circuit 608 ... Channel information acquisition circuit 609-1 ... first interference space candidate calculation circuit 609-2 ... second interference space candidate calculation circuit 610-1 ... first additional interference space selection circuit 610-2 ... Second additional interference space selection circuit 612 ... Transmission rank setting circuit

Claims (16)

A channel response matrix including a plurality of transmitting antenna elements and representing a propagation environment between a receiving antenna and a transmitting antenna element included in a communication partner, or between a receiving beam and a transmitting antenna formed using the receiving antenna. A wireless communication method for estimating, determining a transmission weight suitable for a propagation environment, performing transmission weighting on a transmission signal, and transmitting the transmission signal to a plurality of communication partners,
For each communication partner that performs communication, from the channel response vector group between the reception antenna or reception beam of the communication partner other than the communication partner and the transmission antenna of the transmission device, determining the interference space basis vector group of the communication partner;
Calculating an orthogonal spatial channel response matrix obtained by extracting a component orthogonal to the interference spatial basis vector from the channel response matrix for the communication partner;
Determining a transmission weight obtained by a linear operation from an orthogonal spatial channel response matrix;
Can be obtained as an interference space basis vector candidate when a unit vector obtained by using the orthogonalization method on the row vector group of the complex conjugate matrix of the communication partner's channel response matrix or when the channel response matrix of the communication partner is singularly decomposed Calculating a transmission-side eigenvector or a vector obtained by QR decomposition;
Inner product value of the vectors related to different communication partners
P _{kl, ab} = (ω _{k, a} , ω _{l, b} )
(K ≠ _l, ω _{k, a} is the a th vector for the k th communication partner, and ω _{l, b} is the b th vector for the l th communication partner.)
And a step of selecting a vector such that the inner product value is low and determining an interference space basis vector. A channel response matrix including a plurality of transmitting antenna elements and representing a propagation environment between a receiving antenna and a transmitting antenna element included in a communication partner, or between a receiving beam and a transmitting antenna formed using the receiving antenna. A wireless communication method for estimating, determining a transmission weight suitable for a propagation environment, performing transmission weighting on a transmission signal, and transmitting the transmission signal to a plurality of communication partners,
A step of determining a first communication partner during simultaneous communication;
A step of determining a transmission weight used when performing transmission weighting on a transmission signal in the first communication sequence of the first communication partner as a first transmission weight by linear calculation from a channel response matrix for the first communication partner. When,
In order to determine the transmission weight used when performing transmission weighting on the transmission signals in the communication sequence other than the first communication sequence of the first communication partner and the communication sequence of the second communication partner, for each communication partner, A channel response vector group between the receiving antennas or receiving beams of all the communication partners other than the communication partner whose transmission weight is to be determined and the transmission antenna of the transmitting apparatus, and the first transmission weight, formed in the first communication partner Determining a set of interference space basis vectors for all communicating parties from the channel response vector between the received beam and the transmit antenna;
Calculating an orthogonal spatial channel response matrix obtained by extracting components orthogonal to the interference spatial basis vector from the channel response matrix for all communication partners;
Determining a second transmission weight obtained by a linear operation from an orthogonal spatial channel response matrix;
Can be obtained as an interference space basis vector candidate when a unit vector obtained by using the orthogonalization method on the row vector group of the complex conjugate matrix of the communication partner's channel response matrix or when the channel response matrix of the communication partner is singularly decomposed Calculating a transmission-side eigenvector or a vector obtained by QR decomposition;
Inner product value of the vectors related to different communication partners
P _{kl, ab} = (ω _{k, a} , ω _{l, b} )
(K ≠ _l, ω _{k, a} is the a th vector for the k th communication partner, and ω _{l, b} is the b th vector for the l th communication partner.)
And a step of selecting a vector such that the inner product value is low and determining an interference space basis vector. A channel response matrix including a plurality of transmitting antenna elements and representing a propagation environment between a receiving antenna and a transmitting antenna element included in a communication partner, or between a receiving beam and a transmitting antenna formed using the receiving antenna. A wireless communication method for estimating, determining a transmission weight suitable for a propagation environment, performing transmission weighting on a transmission signal, and transmitting the transmission signal to a plurality of communication partners,
A step of determining a first communication partner during simultaneous communication;
When transmission of the first communication sequence of the first communication partner is performed, interference due to this communication sequence cannot be largely removed or a communication partner that does not have a sufficient decoding device is determined, and these communication partners are determined. Determining as a communication partner capable of interference;
Determining a first interference space basis vector from a channel response matrix for a non-interacting communication partner;
Calculating a first orthogonal spatial channel response matrix obtained by extracting a component orthogonal to the first interference spatial basis vector from the channel response matrix for the first communication partner;
Determining a first transmission weight of a first communication partner obtained by a linear operation from a first orthogonal spatial channel response matrix;
In order to determine the communication weight other than the first communication line of the first communication partner and the transmission weight used for the communication line of the second communication partner, other than the communication partner trying to determine the transmission weight for each communication partner Channel response vector groups between the reception antennas or reception beams of all communication partners of and the transmission antennas of the transmission apparatus, and between the reception beams and transmission antennas formed at the first communication partner with respect to the first transmission weight. Determining interference space basis vectors of all communication partners from the channel response vectors of
Calculating an orthogonal spatial channel response matrix obtained by extracting components orthogonal to the interference spatial basis vector from the channel response matrix for all communication partners;
Determining a second transmission weight obtained by a linear operation from an orthogonal spatial channel response matrix;
Can be obtained as an interference space basis vector candidate when a unit vector obtained by using the orthogonalization method on the row vector group of the complex conjugate matrix of the communication partner's channel response matrix or when the channel response matrix of the communication partner is singularly decomposed Calculating a transmission-side eigenvector or a vector obtained by QR decomposition;
Inner product value of the vectors related to different communication partners
P _{kl, ab} = (ω _{k, a} , ω _{l, b} )
(K ≠ _l, ω _{k, a} is the a th vector for the k th communication partner, and ω _{l, b} is the b th vector for the l th communication partner.)
And a step of selecting a vector such that the inner product value is low and determining an interference space basis vector. A channel response matrix including a plurality of transmitting antenna elements and representing a propagation environment between a receiving antenna and a transmitting antenna element included in a communication partner, or between a receiving beam and a transmitting antenna formed using the receiving antenna. A wireless communication method for estimating, determining a transmission weight suitable for a propagation environment, performing transmission weighting on a transmission signal, and transmitting the transmission signal to a plurality of communication partners,
Determining a plurality of first communication partners from performing simultaneous communication;
When transmission of the first communication sequence of the first communication partner is performed, interference due to this communication sequence cannot be largely removed or a communication partner that does not have a sufficient decoding device is determined, and these communication partners are determined. Determining as a communication partner capable of interference;
Determining a first interference space basis vector from a channel response matrix of a communication partner other than the non- interference communication partner among a plurality of first communication partners and a channel response matrix for the non-interference communication partner;
Calculating a first orthogonal spatial channel response matrix obtained by extracting a component orthogonal to the first interference spatial basis vector from the channel response matrix for the first communication partner;
Determining a first transmission weight of a first communication partner obtained by a linear operation from a first orthogonal spatial channel response matrix;
In order to determine the communication weight other than the first communication line of the first communication partner and the transmission weight used for the communication line of the second communication partner, other than the communication partner trying to determine the transmission weight for each communication partner Channel response vector groups between the reception antennas or reception beams of all communication partners of and the transmission antennas of the transmission apparatus, and between the reception beams and transmission antennas formed at the first communication partner with respect to the first transmission weight. Determining interference space basis vectors of all communication partners from the channel response vectors of
Calculating an orthogonal spatial channel response matrix obtained by extracting components orthogonal to the interference spatial basis vector from the channel response matrix for all communication partners;
Determining a second transmission weight obtained by a linear operation from an orthogonal spatial channel response matrix;
Can be obtained as an interference space basis vector candidate when a unit vector obtained by using the orthogonalization method on the row vector group of the complex conjugate matrix of the communication partner's channel response matrix or when the channel response matrix of the communication partner is singularly decomposed Calculating a transmission-side eigenvector or a vector obtained by QR decomposition;
Inner product value of the vectors related to different communication partners
P _{kl, ab} = (ω _{k, a} , ω _{l, b} )
(K ≠ _l, ω _{k, a} is the a th vector for the k th communication partner, and ω _{l, b} is the b th vector for the l th communication partner.)
And a step of selecting a vector such that the inner product value is low and determining an interference space basis vector. The wireless communication method according to claim 1, wherein
The unit vector obtained by using the orthogonalization method for the row vector group of the complex conjugate matrix of the channel response matrix of the communication partner as the interference space basis vector, or the transmission obtained when singular value decomposition of the channel response matrix of the communication partner Calculating a side eigenvector or a vector obtained by QR decomposition;
A wireless communication method comprising: The wireless communication method according to claim 1, wherein
Can be obtained as an interference space basis vector candidate when a unit vector obtained by using the orthogonalization method on the row vector group of the complex conjugate matrix of the communication partner's channel response matrix or when the channel response matrix of the communication partner is singularly decomposed Calculating a transmission-side eigenvector or a vector obtained by QR decomposition;
Estimating a signal-to-noise ratio of a communication sequence obtained by using the obtained vector as a transmission weight;
Calculating an inner product value of the vectors related to different communication partners;
A step of estimating degradation of a signal-to-noise ratio when these vectors are used from the correlation between vectors and selecting a combination having the highest expected value of transmission capacity when performing communication. Wireless communication method. The wireless communication method according to claim 1, wherein
Can be obtained as an interference space basis vector candidate when a unit vector obtained by using the orthogonalization method on the row vector group of the complex conjugate matrix of the communication partner's channel response matrix or when the channel response matrix of the communication partner is singularly decomposed Calculating a transmission-side eigenvector or a vector obtained by QR decomposition;
Estimating the signal-to-noise ratio of the communication sequence originally obtained by using the obtained vector as a transmission weight;
Calculating an inner product value of the vectors related to different communication partners;
Estimating the degradation of the signal-to-noise ratio when using those vectors from the correlation between the vectors, and selecting a combination that can achieve the maximum transmission rate with acceptable quality in view of the selectable modulation scheme; and A wireless communication method comprising: The wireless communication method according to claim 1, wherein
After determining the transmission weight or the first and second transmission weights, the communication partner corresponding to the determined transmission weight of the communication partner is again used as the interference space basis vector or the first and second interference space basis vectors. Computing a channel response vector between the receive beam and transmit antenna elements of
Calculating an orthogonal spatial channel response matrix, or a first and second orthogonal spatial channel response matrix based on using a newly defined interference space basis vector, or first and second interference space basis vectors;
A step of newly determining transmission weights or first and second transmission weights from the obtained orthogonal spatial channel response matrix or the first and second orthogonal spatial channel response matrices. A wireless communication method characterized by repeating the number of times. A transmitter station with multiple antenna elements;
A MIMO (Multiple Input Multiplex) composed of a plurality of communication partner stations having one or a plurality of antenna elements, and composed of a transmitting station and antenna elements of the plurality of communication partner stations, or beams formed on these antenna elements. Mu communication in a wireless communication system capable of performing MIMO communication by spatially multiplexing one or a plurality of signal sequences at the same frequency channel and the same time to a plurality of communication counterpart stations via an Output channel. In a wireless communication apparatus that transmits signals by L (1) to L (Mu) spatial multiplexing,
The transmission spatial multiplexing number Mu × (L (1) + L (2) +... + L (Mu)) or more, MT (MT> 1: integer) antenna elements,
A radio unit connected to each antenna element, converting a received signal into a baseband signal at the time of reception and outputting it to a channel information acquisition circuit; and transmitting a radio signal from the antenna element as a radio signal at the time of transmission;
A channel information acquisition circuit that estimates a channel response matrix for a communication partner from a signal input from a wireless unit, and outputs a channel response matrix of a communication partner to be transmitted to an interference space arithmetic circuit when a communication partner and a transmission signal are determined; ,
Based on the channel response matrix input from the channel information acquisition circuit, an interference space arithmetic circuit that calculates an interference space basis vector for each communication partner,
Based on the interference space basis vector input from the interference space calculation circuit and the channel response matrix input from the channel response matrix acquisition circuit, an orthogonal spatial channel response matrix that reduces interference with other parties that perform communication is calculated. An orthogonal space arithmetic circuit;
The orthogonal spatial channel response matrix calculated in the orthogonal spatial arithmetic circuit is used as an input signal, a transmission weight obtained by performing a linear operation on the orthogonal spatial channel response matrix is determined, output to the transmission signal conversion circuit, and transmitted to each communication sequence. A transmission weight calculation circuit that outputs a transmission mode consisting of a modulation scheme and a coding rate to be applied to a modulation circuit and a data division circuit;
A transmission weight determination block including the channel response matrix acquisition circuit, interference space calculation circuit, orthogonal space calculation circuit, and transmission weight calculation circuit;
A data division circuit that divides transmission data into the number of communication sequences according to a transmission mode, a modulation circuit that modulates each signal sequence,
A transmission signal conversion circuit for multiplying each modulated communication sequence by a transmission weight determined by a transmission weight arithmetic circuit and outputting to a radio unit connected to a corresponding antenna element;
Can be obtained as an interference space basis vector candidate when a unit vector obtained by using the orthogonalization method on the row vector group of the complex conjugate matrix of the communication partner's channel response matrix or when the channel response matrix of the communication partner is singularly decomposed A calculation unit for calculating a transmission side eigenvector or a vector obtained by QR decomposition;
Inner product value of the vectors related to different communication partners
P _{kl, ab} = (ω _{k, a} , ω _{l, b} )
(K ≠ _l, ω _{k, a} is the a th vector for the k th communication partner, and ω _{l, b} is the b th vector for the l th communication partner.)
And a determination unit that selects a vector so as to reduce the inner product value and determines an interference space basis vector. The wireless communication device according to claim 9 , wherein
The transmission weight determination block includes:
A channel response matrix is estimated from a signal input from the radio unit, and when a communication partner and a transmission signal are determined, a channel information acquisition circuit that outputs the channel response matrix of the communication partner to an interference space candidate calculation circuit;
From the input channel response matrix, as an interference space basis vector candidate as an interference space basis vector candidate, a unit vector obtained by using an orthogonalization method on a row vector group of a complex conjugate matrix of a channel response matrix of a communication partner, An interference space candidate calculation circuit that outputs a transmission side eigenvector obtained by performing singular value decomposition on the channel response matrix of the communication partner, or a vector obtained by linear calculation that can be approximated to this eigenvector, to the communication space selection circuit;
A communication space selection circuit that selects a communication partner that actually performs transmission, the number of communication sequences, and an interference space basis vector used as an interference space;
An orthogonal space calculation that calculates an orthogonal spatial channel response matrix that reduces interference with other parties that communicate with the interference space basis vector input from the communication space selection circuit and the channel response matrix input from the channel information acquisition circuit Circuit,
The orthogonal spatial channel response matrix calculated in the orthogonal spatial arithmetic circuit is used as an input signal, a transmission weight obtained by performing a linear operation on the orthogonal spatial channel response matrix is determined, output to the transmission signal conversion circuit, and transmitted to each communication sequence. A wireless communication apparatus comprising: a transmission circuit that outputs a transmission mode to be applied to a modulation circuit and a data division circuit. The wireless communication device according to claim 9 , wherein
The transmission weight determination block includes:
A channel response matrix is estimated from a signal input from the radio unit, and when a communication partner and a transmission signal are determined, a channel information acquisition circuit that outputs the channel response matrix of the communication partner to an interference space candidate calculation circuit;
From the input channel response matrix, as an interference space basis vector candidate as an interference space basis vector candidate, a unit vector obtained by using an orthogonalization method on a row vector group of a complex conjugate matrix of a channel response matrix of a communication partner, An interference space candidate calculation circuit that outputs a transmission side eigenvector obtained by performing singular value decomposition on the channel response matrix of the communication partner, or a vector obtained by linear calculation that can be approximated to this eigenvector, to the communication space selection circuit;
Select the communication partner that actually performs transmission, the number of communication sequences, and the interference space basis vector to be used as the interference space, and from the candidate interference space basis vectors that were not selected, the channel response matrix is not completely orthogonal. A communication space selection circuit that selects an interference space basis vector and outputs it to an orthogonal space arithmetic circuit;
Orthogonal spatial channel response matrix that reduces interference to non-communicating parties from the interference space basis vectors and quasi-interference space basis vectors input from the communication space selection circuit and the channel response matrix input from the channel information acquisition circuit An orthogonal space arithmetic circuit for calculating
The orthogonal spatial channel response matrix calculated in the orthogonal spatial arithmetic circuit is used as an input signal, a transmission weight obtained by performing a linear operation on the orthogonal spatial channel response matrix is determined, output to the transmission signal conversion circuit, and transmitted to each communication sequence. A wireless communication apparatus comprising: a transmission circuit that outputs a transmission mode to be applied to a modulation circuit and a data division circuit. A wireless communication apparatus according to claim 9,
The transmission weight determination block includes:
A channel response matrix is estimated from a signal input from the radio unit, and when a communication partner and a transmission signal are determined, a channel information acquisition circuit that outputs the channel response matrix of the communication partner to an interference space candidate calculation circuit;
From the input channel response matrix, as an interference space basis vector candidate as an interference space basis vector candidate, a unit vector obtained by using an orthogonalization method on a row vector group of a complex conjugate matrix of a channel response matrix of a communication partner, An interference space candidate calculation circuit that outputs a transmission side eigenvector obtained by performing singular value decomposition on the channel response matrix of the communication partner, or a vector obtained by linear calculation that can be approximated to this eigenvector, to the communication space selection circuit;
Influence of the time variation, the error of the device, a vector it is anticipated that an interference space when actually transmitting, the additional interference space selection circuit for outputting as the additional interference space basis vectors,
Select the communication partner that actually performs transmission, the number of communication sequences, and the interference space basis vector to be used as the interference space, and from the candidate interference space basis vectors that were not selected, the channel response matrix is not completely orthogonal. Select an interference space basis vector, convert the additional interference space basis vector input from the additional interference space selection circuit into a vector orthogonal to the interference space basis vector, add to the interference space basis vector or the quasi-interference space basis vector, A communication space selection circuit that outputs to an orthogonal space arithmetic circuit;
Orthogonal spatial channel response matrix that reduces interference to non-communicating parties from the interference space basis vectors and quasi-interference space basis vectors input from the communication space selection circuit and the channel response matrix input from the channel information acquisition circuit An orthogonal space arithmetic circuit for calculating
The orthogonal spatial channel response matrix calculated in the orthogonal spatial arithmetic circuit is used as an input signal, a transmission weight obtained by performing a linear operation on the orthogonal spatial channel response matrix is determined, output to the transmission signal conversion circuit, and transmitted to each communication sequence. A radio communication apparatus comprising: a modulation circuit and a second transmission weight calculation circuit that outputs a transmission mode to be applied to a data division circuit. The wireless communication device according to claim 9 , wherein
The transmission weight determination block includes:
When the communication partner and the transmission signal are determined, the number of the first communication partner and the number of the first communication series to be preferentially transmitted from the communication partners is determined, and information on the first communication partner and other communication partners is obtained. A transmission rank setting circuit for outputting to the channel information acquisition circuit;
The channel response matrix is estimated from the signal input from the radio unit, and when the communication partner and the transmission signal are input from the transmission rank setting circuit, the channel response matrix of the first communication partner and the number of the first communication series are calculated. A channel information acquisition circuit that outputs a channel response matrix of all communication partners to the second interference space candidate calculation circuit to the first transmission weight calculation circuit;
A first transmission weight calculation circuit that determines a prescribed number of first transmission weights from the input channel response matrix of the first communication partner by linear calculation and outputs the first transmission weight to an interference space candidate calculation circuit;
Using the channel response matrix input from the channel information acquisition circuit, singular value decomposition of the unit vector obtained by using the orthogonalization method for the row vector group of the complex conjugate matrix of the channel response matrix and the channel response matrix of the communication partner A second interference space candidate calculation circuit that outputs to the communication space selection circuit a transmission side eigenvector obtained at the time of the transmission, or a vector obtained by a linear operation that can be approximated to this eigenvector;
An additional interference space selection circuit that outputs, as an additional interference space basis vector, a vector that is expected to become an interference space when actually transmitting due to the effects of time fluctuations and device errors;
Select the communication partner that actually performs transmission, the number of communication sequences, and the interference space basis vector to be used as the interference space, and from the candidate interference space basis vectors that were not selected, the channel response matrix is not completely orthogonal. Select an interference space basis vector, convert the additional interference space basis vector input from the additional interference space selection circuit into a vector orthogonal to the interference space basis vector, add to the interference space basis vector or the quasi-interference space basis vector, A second communication space selection circuit that outputs to a second orthogonal space arithmetic circuit;
Orthogonal space that reduces interference with other parties that communicate with each other from the interference space basis vectors and quasi-interference space basis vectors input from the second communication space selection circuit and the channel response matrix input from the channel information acquisition circuit A second orthogonal space arithmetic circuit for calculating a channel response matrix;
The orthogonal spatial channel response matrix calculated in the second orthogonal spatial operation circuit is used as an input signal, a transmission weight obtained by performing linear operation on the orthogonal spatial channel response matrix is determined, and output to the second transmission signal conversion circuit And a second transmission weight calculation circuit that outputs a transmission mode applied to each communication series to the modulation circuit and the data division circuit. The wireless communication device according to claim 9 , wherein
The transmission weight determination block includes:
When the communication partner and the transmission signal are determined, the number of first communication partners and the number of the first communication series to be transmitted with priority among the communication partners are determined, and the first communication partner and other communication partners are determined. A transmission rank setting circuit that outputs information to a channel information acquisition circuit;
The channel response matrix is estimated from the signal input from the radio unit, and when the communication partner and the transmission signal are input from the transmission rank setting circuit, the channel response matrix of the first communication partner and the number of the first communication series are calculated. A channel information acquisition circuit that outputs to the first orthogonal space arithmetic circuit and the first interference space arithmetic circuit, and outputs channel response matrices of all communication partners to the second interference space candidate arithmetic circuit;
A first interference space calculation circuit that determines a first interference space basis vector of each first communication partner from the input channel response matrix of the first communication partner and outputs the first interference space basis vector to the first orthogonal space calculation circuit; ,
The first transmission weight is defined by linear calculation from the first interference space basis vector input from the first interference space calculation circuit and the channel response matrix of the first communication partner input from the channel information acquisition circuit. A first transmission weight calculation circuit that determines the number and outputs to the interference space candidate calculation circuit;
Using the channel response matrix input from the channel information acquisition circuit, singular value decomposition of the unit vector obtained by using the orthogonalization method for the row vector group of the complex conjugate matrix of the channel response matrix and the channel response matrix of the communication partner A second interference space candidate calculation circuit that outputs to the communication space selection circuit a transmission side eigenvector obtained at the time of the transmission, or a vector obtained by a linear operation that can be approximated to this eigenvector;
An additional interference space selection circuit that outputs, as an additional interference space basis vector, a vector that is expected to become an interference space when actually transmitting due to the effects of time fluctuations and device errors;
Select the communication partner that actually performs transmission, the number of communication sequences, and the interference space basis vector to be used as the interference space, and from the candidate interference space basis vectors that were not selected, the channel response matrix is not completely orthogonal. Select an interference space basis vector, convert the additional interference space basis vector input from the additional interference space selection circuit into a vector orthogonal to the interference space basis vector, add to the interference space basis vector or the quasi-interference space basis vector, A second communication space selection circuit that outputs to a second orthogonal space arithmetic circuit;
Orthogonal space that reduces interference with other parties that communicate with each other from the interference space basis vectors and quasi-interference space basis vectors input from the second communication space selection circuit and the channel response matrix input from the channel information acquisition circuit A second orthogonal space arithmetic circuit for calculating a channel response matrix;
The orthogonal spatial channel response matrix calculated in the second orthogonal spatial arithmetic circuit is used as an input signal, a transmission weight obtained by performing a linear operation on the orthogonal spatial channel response matrix is determined, output to the transmission signal conversion circuit, A radio communication apparatus comprising: a modulation mode and a second transmission weight calculation circuit that outputs a transmission mode applied to a communication sequence to a data division circuit. The wireless communication device according to claim 9 , wherein
The transmission weight determination block includes:
When the communication partner and the transmission signal are determined, the number of first communication partners and the number of the first communication series to be transmitted with priority among the communication partners are determined, and the first communication partner and other communication partners are determined. A transmission rank setting circuit that outputs information to a channel information acquisition circuit;
The channel response matrix is estimated from the signal input from the radio unit, and when the communication partner and the transmission signal are input from the transmission rank setting circuit, the channel response matrix of the first communication partner and the number of the first communication series are calculated. A channel information acquisition circuit that outputs to the first orthogonal space computation circuit and the first interference space candidate computation circuit, and outputs channel response matrices of all communication partners to the second interference space candidate computation circuit;
A first interference space candidate calculation that calculates a first interference space basis vector candidate of each first communication partner from the input channel response matrix of the first communication partner and outputs to the first communication space calculation circuit Circuit,
The first interference space basis vector used as the first interference space is selected from the input first interference space basis vector candidates, and the channel response is selected from the first interference space basis vector candidates that are not selected. A first communication space selection circuit that selects a first quasi-interference space basis vector that is not completely orthogonal to the matrix and outputs the first quasi-interference space basis vector;
Linear calculation is performed from the first interference space basis vector and the first quasi-interference space basis vector input from the first communication space calculation circuit and the channel response matrix of the first communication partner input from the channel information acquisition circuit. A first transmission weight calculation circuit that determines a prescribed number of first transmission weights and outputs the first transmission weight to a second interference space candidate calculation circuit,
Using the channel response matrix input from the channel information acquisition circuit, singular value decomposition of the unit vector obtained by using the orthogonalization method for the row vector group of the complex conjugate matrix of the channel response matrix and the channel response matrix of the communication partner A second interference space candidate calculation circuit that outputs to the communication space selection circuit a transmission side eigenvector obtained at the time of the transmission, or a vector obtained by a linear operation that can be approximated to this eigenvector;
An additional interference space selection circuit that outputs, as an additional interference space basis vector, a vector that is expected to become an interference space when actually transmitting due to the effects of time fluctuations and device errors;
Select the communication partner that actually performs transmission, the number of communication sequences, and the interference space basis vector to be used as the interference space, and from the candidate interference space basis vectors that were not selected, the channel response matrix is not completely orthogonal. interference space basis vectors is selected, the additional interference space basis vectors input from the additional interference space selection circuit, into a vector which is orthogonal to the interference spatial basis vectors, or the interference spatial basis vectors, in addition to the quasi-interference space basis vectors, A second communication space selection circuit that outputs to the orthogonal space arithmetic circuit;
Orthogonal spatial channel response matrix that reduces interference to non-communicating parties from the interference space basis vectors and quasi-interference space basis vectors input from the communication space selection circuit and the channel response matrix input from the channel information acquisition circuit An orthogonal space arithmetic circuit for calculating
The orthogonal spatial channel response matrix calculated in the orthogonal spatial arithmetic circuit is used as an input signal, a transmission weight obtained by performing a linear operation on the orthogonal spatial channel response matrix is determined, output to the transmission signal conversion circuit, and transmitted to each communication sequence. A wireless communication apparatus comprising: a transmission circuit that outputs a transmission mode to be applied to a modulation circuit and a data division circuit. The wireless communication device according to claim 9 , wherein
The transmission weight determination block includes:
When the communication partner and the transmission signal are determined, the number of first communication partners and the number of the first communication series to be transmitted with priority among the communication partners are determined, and the first communication partner and other communication partners are determined. A transmission rank setting circuit that outputs information to a channel information acquisition circuit;
The channel response matrix is estimated from the signal input from the radio unit, and when the communication partner and the transmission signal are input from the transmission rank setting circuit, the channel response matrix of the first communication partner and the number of the first communication series are calculated. The first orthogonal space arithmetic circuit and the first interference space candidate arithmetic circuit are output, and the channel response matrix of all communication partners is output to the second interference space candidate arithmetic circuit, so that the interference due to the first communication series is completely eliminated. A channel information acquisition circuit that outputs at least a part of the channel response matrix of a communication partner that is difficult to remove to the first interference interference selection circuit;
A first interference space candidate calculation that calculates a first interference space basis vector candidate of each first communication partner from the input channel response matrix of the first communication partner and outputs to the first communication space calculation circuit Circuit,
Due to the effects of time fluctuations and device errors, the vector that is expected to become an interference space when actually transmitting, and the interference space basis vector obtained from the channel response matrix of the communication partner that cannot input interference, A first additional interference space selection circuit for outputting as an additional interference space basis vector;
The first interference space basis vector used as the first interference space is selected from the input first interference space basis vector candidates, and the channel response is selected from the first interference space basis vector candidates that are not selected. A first quasi-interference space basis vector that is not completely orthogonal to the matrix is selected, and the first additional interference space basis vector input from the first additional interference space selection circuit is used as the first interference space basis vector. A first communication space selection circuit that converts to an orthogonal vector and outputs to the first orthogonal space arithmetic circuit in addition to the first interference space basis vector or the first quasi-interference space basis vector;
Linear calculation is performed from the first interference space basis vector and the first quasi-interference space basis vector input from the first communication space calculation circuit and the channel response matrix of the first communication partner input from the channel information acquisition circuit. A first transmission weight calculation circuit that determines a prescribed number of first transmission weights and outputs the first transmission weight to a second interference space candidate calculation circuit,
Communicating an interference space basis vector candidate as a second interference space basis vector candidate from the channel response matrix inputted from the channel information acquisition circuit and the first transmission weight inputted from the first transmission weight calculation circuit. A second interference space candidate calculation circuit that outputs to the space selection circuit;
A second additional interference space selection circuit that outputs, as an additional interference space basis vector, a vector that is expected to become an interference space when actually transmitting due to the influence of time fluctuations or device errors;
Select the communication partner that actually performs transmission, the number of communication sequences, and the interference space basis vector to be used as the interference space, and from the candidate interference space basis vectors that were not selected, the channel response matrix is not completely orthogonal. Select an interference space basis vector, convert the additional interference space basis vector input from the additional interference space selection circuit into a vector orthogonal to the interference space basis vector, add to the interference space basis vector or the quasi-interference space basis vector, A communication space selection circuit that outputs to an orthogonal space arithmetic circuit;
Orthogonal spatial channel response matrix that reduces interference to non-communicating parties from the interference space basis vectors and quasi-interference space basis vectors input from the communication space selection circuit and the channel response matrix input from the channel information acquisition circuit An orthogonal space arithmetic circuit for calculating
The orthogonal spatial channel response matrix calculated in the orthogonal spatial arithmetic circuit is used as an input signal, a transmission weight obtained by performing a linear operation on the orthogonal spatial channel response matrix is determined, output to the transmission signal conversion circuit, and transmitted to each communication sequence. A wireless communication apparatus comprising: a transmission circuit that outputs a transmission mode to be applied to a modulation circuit and a data division circuit.

Images