JP5223939B2 - MIMO radio communication method and MIMO radio communication apparatus - Google Patents

Description

  The present invention relates to a MIMO (Multiple-Input Multiple-Output) radio communication system performed between a transmission terminal having a plurality of antennas and a reception terminal having a plurality of antennas, and a MIMO radio communication apparatus for performing communication by the MIMO radio communication system It is about. In particular, in an eigenmode transmission system, which is one of the MIMO communication systems, a MIMO radio communication system and a MIMO radio communication that perform communication at a high transmission rate when the SNR (Signal-to-Noise Ratio) is high. It relates to the device.

  Due to the recent increase in communication demand, the capacity of communication systems is increasing. This trend is also remarkable in wireless communications. For example, even in IEEE 802.11, which defines the standard of wireless LAN by IEEE (Institute of Electrical and Electronics Engineers, Inc.), which promotes standardization of LAN (Local Area Network), communication capacity has been expanded. The original IEEE802.11 communication capacity was 2 Mbps, but it was expanded to a maximum of 11 Mbps in IEEE802.11b, a maximum of 54 Mbps in IEEE802.11a and g, and a specification of up to 600 Mbps in IEEE802.11n scheduled for termination of standardization in 2007 Is scheduled to be completed.

  A technique adopted in IEEE802.11n or the like as a method for realizing a large capacity of wireless communication is MIMO. FIG. 2 schematically shows a MIMO wireless communication system. The transmitting terminal has N transmitting antennas 202, and the receiving terminal has N receiving antennas 203. Signals transmitted from the transmitting antennas 202-1 to 202-N are defined as t1 to tN, and a transmitting antenna signal vector t is defined as in Expression 1.

  Similarly, the signals received by the receiving antennas 203-1 to 203-N are defined as r1 to rN, and the receiving antenna signal vector r is defined as in Expression 2.

  Then, the conversion from t to r can be expressed by the primary conversion of Equation 3.

  The matrix H representing this primary transformation is called a channel matrix. Actually, noise is generated at the same time, so that a noise component n is added as shown in Equation 4.

  The channel matrix can be estimated by the receiver by transmitting a known signal from the transmitting terminal and receiving it by the receiving terminal. This is called channel matrix estimation, and the transmitted known signal is called a training signal. When performing MIMO wireless communication, channel matrix estimation is performed in advance.

  In FIG. 2, the transmission data signals are x1 to xN, the reception data signals are y1 to yN, and the transmission data signal vector x is defined as Equation 5 and the reception data vector y is defined as Equation 6.

  In the transmission antenna weight unit 201 in the transmission terminal, x is converted to t by primary conversion, and in the reception antenna weight unit 204 in the reception terminal, r is converted to y by primary conversion. The simplest method for realizing MIMO wireless communication is a ZF (Zero-Forcing) method in which a transmission antenna weight is a unit matrix and a reception antenna weight is an inverse matrix of H. In the ZF method, the relationship between x and y is expressed by Equation 7.

  In this way, the transmission data signal can be restored in the receiver by canceling H with its inverse matrix. However, noise is amplified by the inverse matrix of H.

  In addition to the above ZF method, there is a method for realizing MIMO wireless communication called an eigenmode transmission method. In this method, first, SVD (Singular Value Decomposition) expressed by Equation 8 is obtained.

  S is a diagonal matrix whose components are all positive real numbers, and U and V are unitary matrices. H on the shoulder of V represents Hermitian conjugate (= transposition + complex conjugate). The diagonal component of S is called the singular value. The first singular value, the second singular value,. . . Let's call them in descending order of singular values. From this result, the transmission antenna weight is V and the reception antenna weight is U Hermitian conjugate. In the eigenmode transmission system, the relationship between x and y is expressed by Equation 9.

  Here, the property that the Hermitian conjugate of the unitary matrix is equal to the inverse matrix is used. Further, due to the nature of the unitary matrix, the magnitude of the noise does not change at all in the term of the last noise n in Equation 9, and considering that S is a diagonal matrix, the received data signal has a singular value in the transmitted data signal. The multiplied signals are obtained separately. Data communication from x1 to y1 is called the first eigenmode, and the second eigenmode, third eigenmode,. . . Call it. In each eigenmode, a singular value squared propagation gain (loss if less than 1) occurs. It is generally known that this eigenmode transmission system is a MIMO wireless communication system that can achieve the largest communication capacity.

  However, unlike the ZF method, the eigenmode transmission method needs to set the transmission antenna weight to a matrix calculated from H. Furthermore, since H is estimated by the receiver, it is necessary to feed back the information of H from the receiving terminal to the transmitting terminal. Therefore, it is necessary to exchange information shown in FIG. Note that both the transmitting terminal and the receiving terminal have a transceiver function. First, the transmitting terminal transmits training data, and the receiving terminal receives it to estimate the channel matrix. The obtained channel matrix is returned to the transmitting terminal, and the transmitting antenna weight is determined by SVD of the channel matrix at the transmitting terminal. Next, the transmitting terminal transmits the training signal after performing transmission antenna weight processing, and the receiving terminal re-estimates the channel matrix from the received training signal to determine the receiving antenna weight. Thereafter, the transmission data is transmitted by the transmission terminal transmitting the transmission data signal subjected to the transmission antenna weight process, and the data is restored by the reception antenna weight at the reception terminal. In FIG. 3, the transmitting antenna weight is determined by SVD at the transmitting terminal, but the transmitting antenna weight can also be obtained at the receiving terminal as shown in FIG. In this case, the transmission antenna weight is fed back.

  In FIGS. 3 and 4, the first SVD does not determine the reception antenna weight, but obtains the reception antenna weight by obtaining a channel matrix using a training signal subjected to transmission antenna weight processing separately. The channel matrix H ′ estimated at this time is expressed by Equation 10.

  If the ZF method is used to determine the receiving antenna weight, the receiving antenna weight R is given by Equation 11.

  Therefore, the relationship between x and y can be expressed by Equation 12.

  That is, in the nth eigenmode, the magnitude of noise changes according to the reciprocal of the nth singular value, and the SNR changes in proportion to the square of the singular value. It is widely known that there are other MMSE (Minimum Mean Square Error) methods and MLD (Maximum Likelihood Detection) methods for determining the receiving antenna weight.

  This method is used for the following reason. The first reason is to cope with the time change of the channel matrix. Even when the channel matrix changes during data transmission, it is possible to set the receiving antenna weight according to the current channel matrix, so that it is possible to suppress degradation of reception characteristics. The second reason is that feedback information can be reduced. In OFDM (Orthogonal Frequency Division Multiplexing) communication such as that used in a wireless LAN, it is necessary to set a transmission antenna weight for each subcarrier, so the amount of feedback information is very large. For this reason, feedback information is appropriately thinned out, so there may be a large difference between the transmission antenna weight and the weight obtained from the SVD of the channel matrix. Since the receiving antenna weight is set in consideration, the characteristic deterioration can be suppressed.

  FIG. 6 shows the probability distribution of the propagation gain calculated from the square of the singular value. Assuming the case of four transmitting antennas and four receiving antennas, each component of the channel matrix is assumed to be a random variable (Rayleigh fading) according to an independent Rayleigh distribution. The propagation loss from each transmitting antenna to the receiving antenna is 0 dB. For reference, SISO (Single-Input Single-Output) with one transmitting antenna and one receiving antenna is also shown. As assumed, SISO has an average propagation gain of 0 dB. On the other hand, in the case of MIMO, the average propagation gain is about 8 dB in the fourth eigenmode, but an average propagation gain close to 10 dB is obtained in the first eigenmode. Therefore, for example, in a communication environment where a reception SNR of 30 dB can be obtained, an SNR of 40 dB can be effectively achieved in the first eigenmode. Therefore, in order to effectively use the first eigenmode, it is important to transmit a large amount of information by adopting a large multi-level modulation.

  However, when modulating with a large multilevel number, high accuracy is required for an RF (Radio Frequency) circuit. Factors that degrade accuracy in the RF circuit include IQ mismatch and non-linearity of the power amplifier. The accuracy of the circuit used in the current wireless LAN is limited to 64QAM modulation, and 256QAM or more is extremely difficult. Therefore, IEEE802.11a, g adopts only four of BPSK, QPSK, 16QAM, and 64QAM as standard, and cannot use 256QAM. Even in IEEE802.11n where MIMO is introduced, 256QAM is not adopted. Therefore, even if a high SNR can be achieved, the modulation multi-level number cannot be increased, and the communication capacity cannot be increased.

  Patent Literature 1 and Non-Patent Literature 1 propose methods for solving the above problems. In this method, in a MIMO-OFDM wireless communication system that performs eigenmode transmission, the same modulation multilevel number is taken for all subcarriers and all eigenmodes, and the output of the error correction encoder is assigned to different subcarriers in order. At the time of allocation, the eigenmode is switched for each subcarrier. In this method, in the eigenmode with a small singular value, the number of modulation multivalues is too large for the singular value, so errors frequently occur. However, in the eigenmode with a large singular value, the modulation multivalue with respect to the singular value is reversed. Errors are less likely to occur because the number is small. For this reason, errors that occur in the eigenmode with a small singular value are corrected by error correction processing, and communication with a large communication capacity becomes possible.

JP 2005-323217 A

Yasuhiko Tanabe, Hiroki Shoki, Hiroshi Tsurumi, "Study of transmission scheme considering nonlinear distortion in MIMO-OFDM", 2005 IEICE General Conference, B-5-79

  The present invention realizes a large communication capacity in a MIMO wireless communication system that performs eigenmode transmission even in a communication environment where a high SNR can be obtained without increasing the number of modulation levels.

  As described above, Patent Document 1 and Non-Patent Document 1 propose a method for solving this problem. However, these methods can only deal with multi-carrier transmission such as OFDM. In addition, the problem is that the modulation multi-value number cannot be increased in the first eigenmode in which a high effective SNR is obtained. However, these methods cannot solve this problem.

  In the MIMO wireless communication system of the present invention, the transmission stream weight is calculated after determining the singular value and transmission antenna weight of each eigenmode by singular value expansion of the channel matrix. Also, at the time of data transmission, conversion processing based on the transmission stream weight calculated here is performed immediately before the transmission antenna weight processing. By this conversion process, an eigenmode with a large singular value and a high effective SNR and an eigenmode with a small singular value and a low effective SNR are combined, and the effective SNR of a mode that requires a large number of modulation multilevels is suppressed. For a mode with a low effective SNR, the SNR is increased.

  A schematic diagram of mode conversion by the transmission stream weight is shown in FIG. In this figure, it is assumed that the number of transmission / reception antennas is three. (A) shows three eigenmodes that can be generated by conventional eigenmode transmission. The effective SNR is expressed by the thickness of the pipe. x1 to 3 are transmission data signals, and y1 to 3 are reception data signals. At this time, it is assumed that the SNR of the eigenmode in which communication is performed from x1 to y1 is high and the optimum modulation multilevel number cannot be selected. In this case, the first eigenmode and the second eigenmode are combined with the transmission stream weight. The state where the transmission stream weight is applied is (b). Then, since x1 and x2 are communicated using both the first eigenmode and the second eigenmode, as shown in (c), both effective SNRs are equal to the SNR of the first eigenmode and the second eigenmode. It becomes an intermediate value of the SNR of the mode. As a result, the SNR state that is too high is improved, so that it is possible to perform communication by selecting the optimum multi-level modulation, and it is possible to increase the communication capacity. The reception stream weight must be applied in addition to the reception antenna weight in accordance with the transmission stream weight application. The channel matrix is estimated from the training signal after the transmission stream weight is applied, and the reception antenna weight is determined from the channel matrix. It can respond.

  If the SNR is high not only in the first eigenmode but also in the modes after the second eigenmode, three or more eigenmodes may be synthesized in descending order of the eigenvalue.

  Since it is necessary to preserve the total transmission power before and after applying the transmission stream weight and to maintain independence between transmission data signals, the transmission stream weight needs to be a unitary matrix. It is convenient to adopt W in Expression 13 as a transmission stream weight when combining the first eigenmode to the nth eigenmode.

  Since all the absolute values of the components of W are equal, all modes can be combined with equal weight.

  When determining the number of eigenmodes to be combined, all eigenmodes whose communication quality indicators for each eigenmode are larger than a preset value and eigenmodes with the largest singular value among the remaining eigenmodes are combined. It ’s fine. Effective SNR can be used as a communication quality index. If the SNR is determined, the communication capacity can be calculated, and the number of modulation multi-levels and the error correction coding rate necessary to achieve the communication capacity are also determined. Therefore, the maximum modulation multi-level that can be adopted in an actual communication system is determined. If the SNR corresponding to the number and the coding rate are calculated backward, it is possible to determine the setting value used for the eigenmode synthesis determination. Similarly, RSSI (Received Signal Strength Indicator) can be used as a communication quality indicator.

  According to the present invention, in a MIMO wireless communication system that performs eigenmode transmission, even in a communication environment where a high SNR is obtained, it is possible to increase the communication capacity by relaxing the limitation due to the upper limit value of the modulation multilevel number.

  FIG. 8 is a graph showing the Shannon communication capacity with respect to the average SNR when the present invention is applied when the number of transmission / reception antennas is 4, the upper limit of the modulation multi-level number is 64QAM, and the error correction coding rate is 3/4. Shown in Propagation was subject to uncorrelated Rayleigh fading. In addition to the method of the present invention, the case of the Zero Forcing method and the case of the conventional eigenmode transmission method are shown. As can be seen from the figure, the method of the present invention can achieve the largest communication capacity. In the vicinity of the average SNR = 25 dB, an improvement of 2.5 dB is obtained as compared with the conventional eigenmode transmission method.

The figure which shows the radio | wireless communication procedure of this invention. Schematic explaining a MIMO wireless communication system. The figure which shows the communication procedure of the conventional eigenmode transmission MIMO communication system. The figure which shows the communication procedure of the conventional eigenmode transmission MIMO communication system. The figure which shows the radio | wireless communication procedure of this invention. The graph which shows the probability distribution of the transmission gain at the time of performing eigenmode MIMO communication with four transmission / reception antennas. FIG. 5 is a schematic diagram illustrating eigenmode synthesis in the wireless communication system of the present invention. The graph which shows the relationship of the communication capacity with respect to average SNR. The functional block diagram of the radio | wireless communication apparatus of this invention. The functional block diagram of the radio | wireless communication apparatus of this invention.

  Embodiments according to the present invention will be described below.

  FIG. 1 is a first embodiment and shows a procedure in the MIMO wireless communication system of the present invention. In this figure, data is transmitted from the transmission terminal to the reception terminal, but both terminals have both transmission and reception functions and can exchange control information and the like.

  First, the transmission terminal transmits training data, and the reception terminal receives training data. This training data is a known signal determined by a standard or the like, and a channel matrix can be estimated by looking at changes in amplitude and phase from the known signal. Next, the receiving terminal returns the channel matrix and the communication quality information obtained by the estimation to the transmitting terminal. SNR and RSSI can be used for the communication quality information. The transmitting terminal receives the returned channel matrix and communication quality information. A singular value expansion is performed on the received channel matrix to obtain a transmission antenna weight in the eigenmode transmission scheme and a singular value of each eigenmode. A communication quality index for each eigenmode is calculated from the obtained singular value and the received communication quality information, and an eigenmode to be synthesized is determined from the communication quality index. The effective SNR and RSSI of each eigenmode can be used as the communication quality index for each eigenmode. The eigenmode to be synthesized can be determined by preliminarily setting an effective SNR or RSSI standard and synthesizing an eigenmode exceeding the standard and the eigenmode having the next largest singular value. Next, a transmission stream weight for combining the determined eigenmodes to be combined is determined. For simplicity, the transmission stream weight can be obtained by Equation (13). Thereafter, training data subjected to transmission stream weight and transmission antenna weight processing is transmitted.

  The receiving terminal receives the training data and estimates the channel matrix. The receiving antenna weight is calculated using the channel matrix. For calculating the receiving antenna weight, the Zero Forcing method, the MMSE method, the MLD method, or the like can be used. Since the estimated channel matrix includes conversion of transmission stream weights and transmission antenna weights, the transmission data signal can be restored by canceling these conversions using the calculated reception antenna weights. Finally, the transmission terminal performs transmission stream weight and transmission antenna weight processing and transmits the transmission data signal, so that the reception terminal can restore the data signal using the reception antenna weight. Communication between the transmitting terminal and the receiving terminal is achieved by the above procedure.

  FIG. 5 shows a procedure in the MIMO wireless communication system of the present invention, which is the second embodiment. As in the first embodiment, data is transmitted from the transmitting terminal to the receiving terminal in this figure, but both terminals have both transmission and reception functions and can exchange control information and the like with each other. it can.

  First, the transmission terminal transmits training data, and the reception terminal receives training data. This training data is a known signal determined by a standard or the like, and a channel matrix can be estimated by looking at changes in amplitude and phase from the known signal. Next, the receiving terminal expands the singular value of the channel matrix obtained by the estimation to obtain the transmission antenna weight in the eigenmode transmission scheme and the singular value of each eigenmode. From the obtained singular value and communication quality information, a communication quality index for each eigenmode is calculated, and an eigenmode to be synthesized is determined from the communication quality index. The effective SNR and RSSI of each eigenmode can be used as the communication quality index for each eigenmode. The eigenmode to be synthesized can be determined by preliminarily setting an effective SNR or RSSI standard and synthesizing an eigenmode exceeding the standard and the eigenmode having the next largest singular value. Next, a transmission stream weight for combining the determined eigenmodes to be combined is determined. For simplicity, the transmission stream weight can be obtained by Equation (13). Then, the determined transmission stream weight and transmission antenna weight are returned to the transmission terminal. Since the transmission stream weight and the transmission antenna weight can be combined by a matrix product, the combined transmission weight may be returned to reduce the amount of information to be returned.

  The transmitting terminal receives the combined transmission weight and transmits training data subjected to transmission weight processing. The receiving terminal receives the training data and estimates the channel matrix. The receiving antenna weight is calculated using the channel matrix. The calculation of the receiving antenna weight can use a Zero Forcing method, an MMSE method, an MLD method, or the like. Since the estimated channel matrix includes transmission weight conversions, the transmission data signal can be restored by canceling these conversions by using the calculated reception antenna weights. Finally, the transmission terminal performs transmission weight processing and transmits a transmission data signal, so that the reception terminal can restore the data signal using the reception antenna weight. Communication between the transmitting terminal and the receiving terminal is achieved by the above procedure.

  FIG. 9 is a functional block diagram of a wireless communication apparatus that performs communication by the MIMO wireless communication system of the present invention, according to the third embodiment.

  The wireless communication apparatus in FIG. 9 has N antennas 101-1 to 101-N and is connected to the switches 102-1 to 102-N. The switch 102 is switched to connect the transmission circuit and the antenna when the wireless communication apparatus performs transmission, and to connect the reception circuit and the antenna when reception is performed. The switch 102 is required in a system that employs a TDD (Time Division Duplex) method such as a wireless LAN, and a system that employs an FDD (Frequency Division Duplex) method often used in mobile phones. Then, a filter called Duplexer is arranged.

  During reception, the antenna 102 and the reception analog RF circuit 103 are connected by the switch 102. In the reception analog RF circuit, down conversion is performed, and the reception signal is converted into a baseband analog signal. The output of the reception analog RF circuit is connected to the AD converter 104, and the baseband analog signal is converted into a digital signal. The output of the AD converter 104 is connected to the FFT processing unit 105. The FFT processing unit 105 decomposes the received signal into OFDM subcarriers. Since the wireless LAN employs OFDM, the FFT processing unit 105 is necessary. However, the FFT processing unit is not necessary in a communication system that performs single carrier transmission. The output of the FFT processing unit 105 is branched into two, and one is connected to the channel matrix estimation unit 110. A channel matrix estimation unit 110 estimates a channel matrix when receiving a training signal. The output of the channel matrix estimator is connected to two blocks. One is input to the reception antenna weight calculation unit 111. The reception antenna weight calculation unit 111 calculates the reception antenna weight from the estimated channel matrix by the Zero Forcing method, the MMSE method, the MLD method, or the like. The other output of the two-branched FFT processing unit 105 and the output of the reception antenna weight calculation unit 111 are input to the reception antenna weight processing unit 106. The reception antenna weight processing unit 106 performs reception data signal restoration using the reception antenna weight calculated by the reception antenna weight calculation unit 111 when reception data signal restoration is necessary. The restored data signal is input to the demodulator 107 where it is converted into bit data. The output of the demodulator 107 is input to the error correction decoding and parallel / serial converter 108, where error correction decoding and parallel / serial conversion are performed. The error correction decoding and the output of the parallel / serial converter 108 are input to the channel information extraction unit 109, and when the received data is a channel matrix or communication quality information with the communication partner, the information is extracted. In the case of other received data, it is delivered to the upper layer. The channel matrix extracted by the channel information extraction unit 109 is input to the singular value expansion processing unit 112 and singular value expansion is performed. The transmission antenna weight determined in this way is delivered to the transmission antenna weight processing unit 117. The singular value is input to the transmission stream weight calculation unit 113. The transmission stream weight calculation unit 113 evaluates the communication characteristic index of each eigenmode using the singular value and determines the transmission stream weight. The transmission stream weight determined here is input to transmission stream weight processing section 118.

  Communication data is delivered from the upper layer to the channel information adding unit 122. Further, the output of the channel matrix estimation unit 110 branched into two is also input to the channel information addition unit 122, and when there is a channel matrix to be transmitted, it is transmitted prior to the communication data. The output of the channel information adding unit 122 is input to the serial / parallel conversion and error correction encoder 121, and serial / parallel conversion and error correction coding are performed. The output of the serial / parallel conversion and error correction encoder 121 is modulated by the modulator 120 and then input to the training signal adding unit 119. The training signal adding unit 119 adds a training signal as necessary and transmits it. The output of the training signal adding unit 119 is processed by the transmission stream weight processing unit 118 and the transmission antenna weight processing unit 117. Since the transmission stream weight processing unit 118 and the transmission antenna weight processing unit 117 are both matrix operations, if the transmission stream weight and the transmission antenna weight are combined in advance by a matrix product, the transmission weight processing can be performed by one process. You can also. The output of transmission antenna weight processing section 117 is converted from an OFDM subcarrier signal to a time domain signal by IFFT processing section 116. In the communication system that performs single carrier transmission, the IFFT processing unit 116 is also unnecessary as in the FFT processing unit 105. The output of the IFFT processing unit 116 is converted into an analog signal by the DA converter 115, up-converted by the transmission analog RF circuit 114, and connected to the switch 102. At the time of transmission, the antenna 102 and the transmission analog RF circuit 114 are connected by the switch 102, and a signal is transmitted.

  Hereinafter, the operation of the wireless communication apparatus of FIG. 9 will be described according to the wireless communication procedure shown in FIG. Both the transmitting terminal and the receiving terminal take the structure of the wireless communication apparatus of FIG. When training data is transmitted from the transmission terminal, a training signal is added by a training signal adding unit 119 and transmitted. At this time, the transmission antenna weight processing unit 117 and the transmission stream weight processing unit 118 do not perform weight processing. Next, the training signal is received at the receiving terminal. At this time, the channel matrix estimating unit 110 estimates the channel matrix, adds the channel information estimated by the channel information adding unit 122, and returns the channel information to the transmitting terminal. The transmission terminal receives this channel information, extracts a channel matrix by the channel information extraction unit 109, and performs singular value expansion by the singular value expansion processing unit 112. The transmission antenna weight is transmitted to the transmission antenna weight processing unit 117 from the result of the singular value expansion, and the transmission stream weight calculation unit 113 calculates the transmission stream weight using the singular value and transmits it to the transmission stream weight processing unit 118. Thereafter, a training signal is added by the training signal adding unit 119 and transmitted to the receiving terminal. At this time, the transmission antenna weight processing unit 117 and the transmission stream weight processing unit 118 perform processing using the set transmission weight. The receiving terminal receives the training signal, and the channel matrix estimation unit 110 estimates the channel matrix. From the estimated matrix, the reception antenna weight calculation unit 111 calculates the reception antenna weight and sets it in the reception antenna weight processing unit 106. Thereafter, the transmission terminal performs transmission weight processing and transmits a transmission data signal, so that the reception terminal can restore the data signal using the reception antenna weight, and communication is established.

  As described above, the wireless communication apparatus of FIG. 9 works, and it is possible to realize a large communication capacity without increasing the modulation multi-level number even in a communication environment where a high SNR is obtained.

  FIG. 10 shows a functional block diagram of a wireless communication apparatus that performs communication by the MIMO wireless communication system of the present invention in the fourth embodiment.

  The configuration of the wireless communication apparatus of FIG. 10 is the same as that of the wireless communication apparatus of FIG. 9 differs from FIG. 9 in the positions of the singular value expansion processing unit 112 and the transmission stream weight calculation unit 113. In this configuration, the reception terminal performs singular value expansion and transmission stream weight calculation.

  The channel matrix estimated by the channel matrix estimation unit 110 is input to the singular value expansion processing unit 112 in addition to the reception antenna weight calculation unit 111. The transmission antenna weight obtained by the singular value expansion processing unit 112 is delivered to the channel information adding unit 122. The singular value is input to the transmission stream weight calculation unit 117, where the transmission stream weight is determined and delivered to the channel information addition unit 122. The channel information adding unit 122 has a function of returning the transmission weight to the transmission terminal. However, since the transmission stream weight and the transmission antenna weight can be combined by a matrix product, the combined transmission is used to reduce the amount of information to be returned. You can reply the weight.

  The channel information extraction unit 109 extracts the transmission antenna weight and the transmission stream weight when the transmission weight is returned, and sets the transmission antenna weight processing unit 117 and the transmission stream weight processing unit 118. However, as described above, when a transmission weight in which the transmission antenna weight and the transmission stream weight are combined by a matrix product is returned, it is only necessary to have one processing unit that performs transmission weight processing using the combined transmission weight.

  Hereinafter, the operation of the wireless communication apparatus of FIG. 10 will be described according to the wireless communication procedure shown in FIG. Both the transmitting terminal and the receiving terminal take the structure of the wireless communication apparatus of FIG. When training data is transmitted from the transmission terminal, a training signal is added by a training signal adding unit 119 and transmitted. At this time, the transmission antenna weight processing unit 117 and the transmission stream weight processing unit 118 do not perform weight processing. Next, the training signal is received at the receiving terminal. At this time, the channel matrix estimation unit 110 estimates the channel matrix, and the singular value expansion processing unit 112 performs singular value expansion. The transmission antenna weight is passed to the channel information adding unit 122 from the result of the singular value expansion, and the transmission stream weight calculation unit 113 evaluates the communication characteristic index of each eigenmode using the singular value, and calculates the transmission stream weight to calculate the channel. Delivered to the information adding unit 122. This time, the channel information adding unit 122 adds the transmission weight data and sends it back to the transmitting terminal. The transmission terminal receives the transmission weight data, extracts the transmission weight by the channel information extraction unit 109, and sets it in the transmission antenna weight processing unit 117 and the transmission stream weight calculation unit 113. Thereafter, a training signal is added by the training signal adding unit 119 and transmitted to the receiving terminal. At this time, the transmission antenna weight processing unit 117 and the transmission stream weight processing unit 118 perform processing using the set transmission weight. The receiving terminal receives the training signal, and the channel matrix estimation unit 110 estimates the channel matrix. From the estimated matrix, the reception antenna weight calculation unit 111 calculates the reception antenna weight and sets it in the reception antenna weight processing unit 106. Thereafter, the transmission terminal performs transmission weight processing and transmits a transmission data signal, so that the reception terminal can restore the data signal using the reception antenna weight, and communication is established.

  As described above, it is possible to realize a large communication capacity without increasing the modulation multi-level number even when the wireless communication apparatus of FIG. 10 works or in a communication environment where a high SNR is obtained.

  101 antenna, 102 switch, 103 reception analog RF circuit, 104 AD converter, 105 FFT processing unit, 106 reception antenna weight processing unit, 107 demodulator, 108 error correction decoding and parallel / serial converter, 109 channel information extraction unit, 110 channel matrix estimation unit, 111 reception antenna weight calculation unit, 112 singular value expansion processing unit, 113 transmission stream weight calculation unit, 114 transmission analog RF circuit, 115 DA converter, 116 IFFT processing unit, 117 transmission antenna weight processing unit, 118 transmission stream weight processing unit, 119 training signal addition unit, 120 modulator, 121 serial / parallel conversion and error correction encoder, 122 channel information addition unit, 201 transmission antenna weight processing unit 202 transmitting antenna 203 receiving antenna 204 receiving antenna weight processing unit.

Claims (5)

A MIMO (Multiple-Input Multiple-Output) wireless communication method in which a plurality of data is transmitted between a transmitting station having a plurality of antennas and a receiving station having a plurality of antennas,
A channel matrix in a propagation path between the transmitting station and the receiving station is obtained based on a training signal reception state in which a training signal transmitted from the transmitting station is received at the receiving station , A first step of forming a plurality of modes of the effective propagation path of data ;
A second step of evaluating a communication characteristic index for each of the plurality of modes;
Have a, a third step of performing data communication from the transmission station by combining at least a portion of said plurality of modes to the receiving station,
The third step includes one or more modes in which the communication characteristic index evaluated in the second step is larger than a predetermined reference, and a mode having the largest communication characteristic index among modes smaller than the reference. A MIMO wireless communication method characterized by combining the above . The MIMO wireless communication method according to claim 1, wherein
A MIMO wireless communication method , wherein an effective propagation gain of the mode is used as the communication characteristic index . The MIMO wireless communication method according to claim 1, wherein
In the third step, according to the communication quality of each mode after the synthesis of the mode, you characterized by adaptively controlling the coding rate of the modulation multi-level number and the error correction code for each mode M IMO wireless communication method. A transmitting station in a MIMO (Multiple-Input Multiple-Output) wireless communication system in which a plurality of data is transmitted between a transmitting station having a plurality of antennas and a receiving station having a plurality of antennas,
Based on a channel matrix in a propagation path between the transmitting station and the receiving station, a mode forming unit that forms a plurality of modes of the effective propagation path of the plurality of data;
A stream weight for combining one or more modes having a communication characteristic index larger than a predetermined reference and a mode having the largest communication characteristic index among modes smaller than the reference is calculated, and the mode is determined by multiplying the stream weight. A stream weight processing unit to be combined;
And a radio communication unit that transmits a transmission signal to which the stream weight is applied from a plurality of antennas. A receiving station in a MIMO (Multiple-Input Multiple-Output) wireless communication system in which a plurality of data is transmitted between a transmitting station having a plurality of antennas and a receiving station having a plurality of antennas,
A channel matrix estimation unit that receives a training signal transmitted from the transmission station and estimates a channel matrix of a propagation path to the transmission station;
A plurality of modes of an effective propagation path of the plurality of data are determined by the channel matrix, and one or more modes whose communication characteristic index is larger than a predetermined reference and the communication characteristic among the modes smaller than the reference A stream weight calculation unit for calculating a stream weight for combining the mode having the largest index;
A transmission unit for transmitting the stream weight to the transmission station,
The receiving station, wherein the stream weight is used for mode transmission from the transmitting station.

Images