JP4765336B2 - Wireless communication apparatus and wireless communication method - Google Patents

Description

  The present invention is a communication (MIMO (Multi Input Multi Output) communication) in which a transmitter having a plurality of antennas and a receiver having a plurality of antennas are paired to form a plurality of logical channels using spatial multiplexing. In particular, the wireless communication device and the wireless communication method for expanding the transmission capacity by using the singular value decomposition (SVD) of the channel matrix H whose elements are the channels corresponding to the transmission / reception antenna pairs. The present invention relates to a wireless communication apparatus and a wireless communication method for performing closed-loop type MIMO transmission for obtaining a weight vector and performing spatial multiplexing and spatial separation of communication data.

  More particularly, the present invention relates to a radio communication apparatus and radio communication method for performing SVD-MIMO communication by applying an OFDM modulation scheme in which a plurality of subcarriers orthogonal to each other are multiplexed and, in particular, arranged on a frequency axis. The present invention also relates to a wireless communication apparatus and a wireless communication method for performing OFDM_MIMO communication while maintaining continuity of frequency characteristics of a plurality of subcarriers.

  A wireless network is attracting attention as a system free from wiring in the conventional wired communication system. As a standard for a wireless network, IEEE (The Institute of Electrical and Electronics Engineers) 802.11 or the like can be cited.

  When a wireless network is constructed indoors, a multipath environment is formed in which a receiving device receives a superposition of a direct wave and a plurality of reflected waves / delayed waves. Multipath causes delay distortion (or frequency selective fading), causes an error in communication, and causes intersymbol interference due to delay distortion. As a main countermeasure against delay distortion, there is a multi-carrier transmission method. In the multi-carrier transmission method, transmission data is distributed and transmitted to a plurality of carriers having different frequencies, so that the band of each carrier becomes narrow and is not easily affected by frequency selective fading.

  For example, in the OFDM (Orthogonal Frequency Division Multiplexing) system, which is one of the multicarrier transmission systems, the frequency of each carrier is set such that the subcarriers are orthogonal to each other within a symbol interval. When transmitting information, serial / parallel conversion of the serially sent information is performed for each symbol period slower than the information transmission rate, and a plurality of output data is assigned to each subcarrier, and amplitude and phase modulation is performed for each subcarrier. Then, by performing inverse FFT on the plurality of subcarriers, the subcarriers are converted to a time axis signal and transmitted while maintaining the orthogonality of each subcarrier on the frequency axis. At the time of reception, the reverse operation, that is, FFT is performed to convert the time-axis signal into the frequency-axis signal, and each subcarrier is demodulated according to the modulation method, and parallel / serial converted to the original. The information sent with the serial signal is reproduced.

  That subcarriers are orthogonal to each other means that the peak point of the spectrum of an arbitrary subcarrier always coincides with the zero point of the spectrum of another subcarrier. According to such an OFDM modulation system, the frequency utilization efficiency is very high and it is strong against frequency selective fading interference. The OFDM modulation scheme is adopted as a standard for wireless LAN in IEEE 802.11a / g, for example.

  In addition, since the multi-carrier transmission scheme has a configuration in which a plurality of subcarriers are continuously arranged on the frequency axis, the communication quality is improved by interpolating using the continuity of the frequency characteristics of adjacent subcarriers. The usage form of improving is also conceivable.

  By the way, although the IEEE 802.11a standard supports a modulation scheme that achieves a communication speed of 54 Mbps at the maximum, a wireless standard capable of realizing a higher bit rate is required. For example, in IEEE802.11n, the next-generation wireless LAN standard is formulated with the aim of developing a high-speed wireless LAN technology with an effective throughput exceeding 100 MBPS.

  MIMO (Multi-Input Multi-Output) communication is attracting attention as one of the technologies for realizing high-speed wireless communication. This is a communication system that includes a plurality of antenna elements on both the transmitter side and the receiver side, and realizes a spatially multiplexed transmission path (hereinafter also referred to as “MIMO channel”), which increases the frequency band. In addition, the transmission capacity can be increased according to the number of antennas, and the communication speed can be improved. MIMO communication is a communication method using channel characteristics, and is different from a simple transmission / reception adaptive array. For example, IEEE 802.11a / n employs an OFDM_MIMO scheme that uses OFDM for primary modulation.

  FIG. 8 conceptually shows the MIMO communication system. As shown in the figure, each transceiver is equipped with a plurality of antennas. On the transmission side, a signal obtained by multiplexing a plurality of transmission data by space / time code is distributed to M antennas and transmitted to a plurality of MIMO channels. On the receiving side, the received data can be obtained by space / time decoding the received signal received by the N antennas via the channel. The channel model in this case is composed of a radio wave environment (transfer function) around the transmitter, a channel space structure (transfer function), and a radio wave environment (transfer function) around the receiver. When signals transmitted from each antenna are multiplexed, crosstalk occurs, but each signal multiplexed by signal processing on the receiving side can be correctly extracted without crosstalk.

The number of MIMO channels obtained in a MIMO communication system generally corresponds to the smaller min [M, n] of the number M of transmission antennas and the number N of reception antennas. The antenna weighting coefficient matrix V on the transmission side is constituted by a transmit vector v _i for the number of MIMO channel _{(V = [v 1, v} 2, ..., v min [M, N]). Further, the number of elements of each transmission vector v _i is the number M of transmission antennas.

  Various schemes have been proposed as a configuration method for MIMO transmission. However, whether to exchange channel information between transmission and reception according to the antenna configuration is a major issue in implementation.

  In order to exchange channel information, it is easy to transmit known information (preamble information) only from the transmission side to the reception side. In this case, the transmitter and receiver perform spatial multiplexing transmission independently of each other. This is called an open-loop type MIMO transmission system. Further, as a developed form of this method, there is a closed-loop type MIMO transmission system that creates an ideal spatial orthogonal channel between transmission and reception by feeding back preamble information from the reception side to the transmission side.

  As an open-loop type MIMO transmission system, for example, a V-BLAST (Vertical Bell Laboratories Layered Space Time) system can be cited (for example, see Patent Document 1). On the transmission side, the antenna weight coefficient matrix is not particularly given, and signals are simply multiplexed and transmitted for each antenna. In other words, any feedback procedure for obtaining the antenna weighting coefficient matrix is omitted. The transmitter inserts a training signal for channel estimation on the receiver side, for example, for each antenna in a time division manner before transmitting the multiplexed signal. On the other hand, in the receiver, the channel estimation unit uses the training signal to perform channel estimation, and calculates a channel information matrix H corresponding to each antenna pair. Then, by skillfully combining zero-forcing and canceling, the SN ratio is improved by utilizing the degree of freedom of the antenna generated by canceling, and the decoding accuracy is increased.

Also, as one of the ideal forms of closed-loop type MIMO transmission, there is known an SVD-MIMO scheme that uses singular value decomposition (SVD) of a propagation path function (for example, non-patent literature). 1). In SVD-MIMO transmission, a numerical matrix having channel information corresponding to each antenna pair as an element, that is, a channel information matrix H, is singularly decomposed to obtain UDV ^H and V is given as an antenna weighting coefficient matrix on the transmission side, and reception is performed. U ^H is given as the antenna weighting coefficient matrix on the side. Thus, each MIMO channel is represented as a diagonal matrix D having the square root of each singular value λ _{i as} a diagonal element, and signals can be multiplexed and transmitted without any crosstalk. In this case, a plurality of logically independent transmission paths that are spatially divided, that is, spatially orthogonally multiplexed, can be realized on both the transmitter side and the receiver side. According to the SVD-MIMO transmission method, the maximum communication capacity can theoretically be achieved. For example, if the transceiver has two antennas, the maximum transmission capacity can be doubled.

FIG. 9 conceptually shows the SVD-MIMO transmission system. In SVD-MIMO transmission, a numerical matrix having channel information corresponding to each antenna pair as an element, that is, a channel information matrix H, is singularly decomposed to obtain UDV ^H and V is given as an antenna weighting coefficient matrix on the transmission side, and reception is performed. U ^H is given as the antenna weighting coefficient matrix on the side. Accordingly, each MIMO channel is represented as a diagonal matrix D having the square root of each eigenvalue λ _{i as} a diagonal element, and signals can be multiplexed and transmitted without any crosstalk. In this case, a plurality of logically independent transmission paths that are spatially divided, that is, spatially orthogonally multiplexed, can be realized on both the transmitter side and the receiver side.

  According to the SVD-MIMO transmission method, the maximum communication capacity can theoretically be achieved. For example, if the transceiver has two antennas, the maximum transmission capacity can be doubled.

Here, the mechanism of the SVD-MIMO transmission scheme will be described in detail. If the number of antennas of the transmitter is M, the transmission signal x is represented by an M × 1 vector, and if the number of antennas of the receiver is N, the received signal y is represented by an N × 1 vector. In this case, the channel characteristic is expressed as an N × M numerical matrix, that is, a channel matrix H. The element h _ij of the channel matrix H is a transfer function from the jth transmit antenna to the ith receive antenna. The received signal vector y is expressed by multiplying the transmission signal vector by the channel information matrix and further adding the noise vector n as shown in the following equation (1).

  As described above, when the channel information matrix H is subjected to singular value decomposition, the following equation (2) is obtained.

  Here, the antenna weight coefficient matrix V on the transmission side and the antenna weight matrix U on the reception side are unitary matrices that satisfy the following expressions (3) and (4), respectively.

In other words, the antenna weight matrix U ^H on the reception side is arranged with the normalized eigenvectors of HH ^H , and the antenna weight matrix V on the transmission side is arranged with the normalized eigenvectors of H ^H H arranged. D is a diagonal matrix having the square root of the eigenvalue of H ^H H or HH ^{H as} a diagonal component. The size is a small number out of the number M of transmitting antennas and the number N of receiving antennas, is a square matrix having a size of min (M, N), and is a diagonal matrix.

In the above description, the singular value decomposition using real numbers has been described. U and V are matrices composed of eigenvectors. However, even when an operation that normalizes the eigenvector to 1, that is, normalization, is performed, the number of eigenvectors having different phases does not become single. Exists. Depending on the phase relationship between U and V, the above equation (2) may not hold. That is, U and V are correct, but the phase is arbitrarily rotated. In order to make the phases completely coincide with each other, V is obtained as an eigenvector of H ^H H as usual, and U is obtained by multiplying both sides of the above equation (2) by V from the right to obtain the following equation. .

On the transmitting side, weighting is performed using the antenna weighting coefficient matrix V, and on the receiving side, when weighting is performed using the antenna weighting coefficient matrix U ^H , U and V are unitary matrices (U is N × min ( M, N) and V are M × min (M, N)), as shown in the following equation.

  Here, the received signal y and the transmitted signal x are (min (M, N) × 1) vectors, not vectors determined by the number of transmitting antennas and receiving antennas.

  Since D is a diagonal matrix, each transmission signal can be received without crosstalk. Since the amplitude of each independent MIMO channel is proportional to the square root of the eigenvalue λ, the power of each MIMO channel is proportional to λ.

Since the noise component n is also an eigenvector whose norm is normalized to 1 in the U column, U ^H n does not change its noise power. As the size, U ^H n is a (min (M, N)) vector, which is the same size as y and x.

  Thus, in SVD-MIMO transmission, it is possible to obtain a plurality of logically independent MIMO channels having no crosstalk while having the same frequency and the same time. That is, it is possible to transmit a plurality of data by wireless communication using the same frequency at the same time, and an improvement in transmission speed can be realized.

  Here, points that must be considered when configuring an actual SVD-MIMO transmission / reception system will be described.

In the basic form of the SVD-MIMO transmission method, the receiver performs singular value decomposition on the acquired channel matrix H to obtain a reception weight vector U ^H and a transmission weight vector V used by the transmitter. Is fed back to the transmitter. The transmitter uses this V as a transmission weight.

  However, since the information amount of the transmission weight matrix V fed back to the transmitter side is large, when the V information is thinned out and transmitted, the orthogonal state between the MIMO channels is different due to an error from the true V information. It breaks and crosstalk occurs.

  Therefore, normally, after the transmission weight matrix V acquired on the receiver side is fed back to the transmitter side, the transmitter weights and transmits the reference signal using the matrix V, and the receiver side again determines the channel matrix. get. If the channel matrix is H, the receiver can obtain a channel matrix of HV from the reference signal weighted with V and transmitted.

On the receiver side, an inverse matrix of this HV is obtained and used as a receiving weight. Since H = UDV ^H , HV is as follows.

This is done by using the same U ^H as the normal SVD-MIMO for reception weight, and then multiplying the stream of each separated MIMO channel by a constant obtained from each diagonal element λ _{i of the} diagonal matrix D. is there.

  The configuration in which the matrix V is used as the transmission weight on the transmission side and the reception weight is used on the inverse side of the HV on the receiver side is the same as the performance of normal SVD-MIMO, There is no discrepancy between V on the receiver side and the receiver side. Therefore, such a configuration can be employed in practice.

  In the SVD-MIMO communication system, for example, from the transmitter side, before the user data is spatially multiplexed and transmitted, a reference signal weighted with HV is time-division multiplexed to obtain a reception weight for each MIMO channel. Sent. On the receiver side, HV is acquired based on the reference signal, and its generalized inverse matrix is used for spatial separation.

  As described above, the multicarrier transmission method represented by OFDM has a configuration in which a plurality of subcarriers are continuously arranged on the frequency axis, and therefore uses the continuity of frequency characteristics in adjacent subcarriers. Communication quality can be improved.

However, in OFDM_MIMO communication in which OFDM modulation is applied to spatially multiplexed MIMO communication, transmission is performed on the transmitter side in order to perform transmission / reception weights for each subcarrier by singular value decomposition of the channel matrix H into UDV ^H. When the data is multiplied by the transmission weight matrix V, the continuity in frequency characteristics between the subcarriers is lost. In such a case, the advantage in the multicarrier transmission system that interpolation is performed using the continuity between subcarriers cannot be utilized.

The channel matrix H and the diagonal matrix D obtained by singular value decomposition are unique. On the other hand, there are an infinite number of combinations of the matrices U and V depending on the amount of phase rotation. Since it is arbitrary which transmission weight vector V is used on the transmitter side, the continuity of frequency characteristics between subcarriers cannot be maintained.
Japanese Patent Laid-Open No. 10-84324 http: // radio3. ee. uec. ac. jp / MIMO (IEICE_TS) .pdf (as of October 24, 2003)

  An object of the present invention is to provide a closed-loop type that performs spatial multiplexing and spatial separation of communication data by obtaining weights of transmission and reception by using singular value decomposition of a channel matrix H whose elements are channels corresponding to transmission and reception antenna pairs. It is an object of the present invention to provide an excellent wireless communication apparatus and wireless communication method capable of suitably performing MIMO transmission.

  A further object of the present invention is to provide an excellent radio communication apparatus and radio communication method capable of suitably performing SVD-MIMO communication by applying an OFDM modulation scheme that multiplex-transmits a plurality of subcarriers orthogonal to each other. It is in.

  A further object of the present invention is to provide an excellent radio communication apparatus and radio communication method capable of performing OFDM_MIMO communication while maintaining continuity of frequency characteristics of a plurality of subcarriers arranged on the frequency axis. is there.

The present invention has been made in view of the above problems, and is a multicarrier transmission of spatially multiplexed signals on a plurality of spatial channels that are formed in pairs with a communication device having a plurality of antennas and having a plurality of antennas. A wireless communication device for performing
Channel matrix acquisition means for acquiring a channel matrix H based on a received signal;
Singular value decomposition means for decomposing the channel matrix H into UDV ^H to obtain transmission weight vectors and reception weight vectors in each spatial channel;
Transmission means for spatially multiplexing transmission data using the transmission weight vector and performing multicarrier transmission;
Receiving means for spatially separating a multicarrier signal received using the reception weight vector into signals for each spatial channel, and receiving processing;
The singular value decomposition means performs a rearrangement operation on the transmission weight vector obtained for each subcarrier so as to maintain a predetermined continuity between the subcarriers.
This is a wireless communication device.

  The present invention relates to a MIMO communication system. In this case, one or both of the transmitter and the receiver are configured so that a transmitter having a plurality of antennas and a receiver having a plurality of antennas are paired to form a plurality of independent logical channels, that is, MIMO channels. Communication is performed by spatially multiplexing transmission signals by combining antennas. According to the MIMO communication system, it is possible to increase the transmission capacity according to the number of antennas without increasing the frequency band, thereby achieving an improvement in communication speed. In particular, the present invention relates to an SVD-MIMO scheme that obtains transmission / reception weights by singular value decomposition (SVD) of a channel matrix.

  In the present invention, the OFDM modulation scheme is applied in order to improve the frequency utilization efficiency and solve the problem of delay distortion in a multipath environment. A multi-carrier transmission system represented by OFDM has a configuration in which a plurality of subcarriers are continuously arranged on the frequency axis. Therefore, interpolation is performed using the continuity of frequency characteristics in adjacent subcarriers. Communication quality can be improved.

However, in OFDM_MIMO communication in which OFDM modulation is applied to spatially multiplexed MIMO communication, transmission is performed on the transmitter side in order to perform transmission / reception weights for each subcarrier by singular value decomposition of the channel matrix H into UDV ^H. There is a problem in that continuity in frequency characteristics between subcarriers is lost when data is multiplied by a transmission weight matrix V.

The process of singular value decomposition includes an operation of rearranging these singular values (in descending order). If the transmission weight vectors V ₁ , V _{2, etc.} are mixed during this rearrangement operation, the continuity of the frequency characteristics is hindered. Even if the transmission weight vectors V ₁ , V ₂ ... Are mistaken, the communication quality of individual subcarriers is not greatly affected. However, the phase of the transmission weight vector changes due to the mistakes, so the continuity of the frequency characteristics between the subcarriers. Can not expect.

  Therefore, in the present invention, the singular value decomposition means performs a rearrangement operation on the transmission weight vector obtained for each subcarrier so as to maintain a predetermined continuity between the subcarriers. That is, the rearrangement operation is performed on the transmission weight vector of each spatial channel obtained for each subcarrier so as to maintain the continuity of the frequency characteristics between the subcarriers.

  Adjacent subcarriers have higher transmission weight vector correlation. By paying attention to this property and rearranging that the next subcarrier has a highly correlated transmission weight vector, it is possible to eliminate transmission weight vector confusion due to the influence of noise or the like. Specifically, the singular value decomposition means obtains a transmission weight vector for a specific subcarrier in each spatial channel. For other subcarriers, after performing singular value decomposition independently, the transmission weight vector rearrangement operation based on the correlation with the transmission weight vector determined in the adjacent subcarrier on the frequency axis is sequentially performed. I will go.

  However, such a rearrangement operation need not always be activated. For example, the singular value decomposition means monitors the singular value for each subcarrier, and performs correlation calculation and rearrangement processing of transmission weight vectors when the singular values of adjacent subcarriers approach a threshold value or less. By doing so, the processing load can be reduced.

  Further, as one of methods for efficiently performing rearrangement based on correlation of transmission vectors when implemented in hardware, the singular value decomposition means obtains the correlation between adjacent subcarriers in parallel, and It is also possible to match all subcarriers after determining the relationship between matching subcarriers first.

Further, the singular value decomposition means normalizes the phases of the transmission weight vectors V ₁ , V ₂ ... (Or U ₁ , U ₂ ...) And aligns the reference phase, thereby subcarriers arranged on the frequency axis. You may make it compensate the continuity of the frequency characteristic in. The normalization of the phase of the transmission weight vector referred to here corresponds to aligning specific elements so as to have the same phase in each transmission weight vector. For example, when the elements of a vector are expressed in complex numbers, the phases of all the elements are rotated so that the same element of each vector is a positive constant.

  In a MIMO communication system, a transmitter that transmits a packet adds a reference signal weighted with a transmission weight vector for each spatial channel to a transmission packet in order to obtain a reception weight vector on the receiver side. In general, the reference signal for each spatial channel is transmitted by time division multiplexing. However, these reference signal portions become overhead, and if this area is large, the throughput is reduced.

  Therefore, the reference signal region may be sent from the transmitter at intervals of one subcarrier, and the missing portion between them may be interpolated on the receiver side to reduce the reference signal region. That is, on the transmitter side, each subcarrier of the reference signals arranged on the frequency axis is weighted by alternately using the transmission weight vector for each spatial channel in the subcarrier. On the other hand, on the receiver side, the spatial channel reception weight vector that could not be obtained in a certain subcarrier is interpolated based on the spatial channel reception weight vector obtained in the adjacent subcarrier on the frequency axis. To do. If the continuity of the frequency characteristics between the subcarriers arranged on the frequency axis is guaranteed, it is considered that highly accurate interpolation can be performed.

  According to the present invention, it is possible to provide an excellent radio communication apparatus and radio communication method that can suitably perform SVD-MIMO communication by applying an OFDM modulation scheme that multiplex-transmits a plurality of subcarriers orthogonal to each other. it can.

  The present invention also provides an excellent radio communication apparatus and radio communication method capable of performing OFDM_MIMO communication while maintaining continuity in frequency characteristics of a plurality of subcarriers arranged on the frequency axis. be able to.

  Other objects, features, and advantages of the present invention will become apparent from more detailed description based on embodiments of the present invention described later and the accompanying drawings.

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

  The present invention relates to a MIMO communication system that performs communication by spatially multiplexing transmission signals. In a MIMO communication system, a transmitter or receiver is configured such that a transmitter having a plurality of antennas and a receiver having a plurality of antennas are paired to form a plurality of independent logical channels, that is, “MIMO channels”. Antenna synthesis is performed on one or both. According to the MIMO communication system, a large amount of data transmission is realized by consolidating a plurality of RF transmission / reception units into one wireless device. In particular, the present invention relates to an SVD-MIMO scheme that obtains transmission / reception weights by singular value decomposition (SVD) of a channel matrix. In addition, an OFDM modulation scheme is applied in order to increase frequency utilization efficiency and solve the problem of delay distortion in a multipath environment.

  FIG. 1 schematically shows the configuration of a wireless communication apparatus according to an embodiment of the present invention. The illustrated wireless communication apparatus 100 includes a plurality of antenna elements on both the transmitter side and the receiver side, and can operate as a MIMO communication apparatus.

  Each transmission / reception antenna 11a and 11b is connected in parallel with a transmission system and a reception system via switches 12a and 12b, respectively, and wirelessly transmits signals to other wireless communication devices on a predetermined frequency channel, or Collects signals sent from other wireless communication devices. However, the switches 12a and 12b connect the transmission / reception antennas 11a and 11b exclusively to one of the transmission system and the reception system, and cannot perform transmission and reception in parallel.

  Each transmission system includes a modulation coding unit 21, a spatial multiplexing unit 22, an IFFT 23, a preamble / reference adding unit 24, a D / A converter 25 for each antenna, and an analog processing unit 26 for transmission.

  The modulation encoding unit 21 encodes transmission data transmitted from an upper layer of the communication protocol with an error correction code, and maps a transmission signal on a signal space by a predetermined modulation method such as BPSK, QPSK, or 16QAM. . Furthermore, a plurality of MIMO channels are obtained by spatial multiplexing by multiplying the encoded transmission signal by a predetermined transmission weight matrix. At this point, a known data sequence may be inserted as a pilot symbol into the modulation symbol sequence according to the pilot symbol insertion pattern and timing. A pilot signal having a known pattern is inserted for each subcarrier or at intervals of several subcarriers.

  The spatial multiplexing unit 22 spatially multiplexes the data for each MIMO channel by multiplying the transmission weight matrix V obtained by singular value decomposition (SVD) of the channel matrix H acquired based on the received signal from the transmission partner. To do. The acquisition of the channel matrix H and the calculation of the transmission weight matrix V are performed in the channel matrix acquisition unit 40, and details thereof will be described later.

  Strictly speaking, the channel matrix H between the transceivers is different in each direction of the uplink and the downlink, but by performing a transfer function calibration process possessed by the transceiver analog circuit of each transceiver, the channel matrix H is bidirectional. An available channel matrix can be obtained. However, the calibration process itself is not directly related to the gist of the present invention and will not be further described here.

  In IFFT 23, the modulated serial signal is converted into parallel data corresponding to the number of parallel carriers in accordance with the number of parallel carriers and the timing, and then the inverse Fourier transform for the FFT size is performed in accordance with a predetermined FFT size and timing. Do.

  Here, in order to remove intersymbol interference, guard interval sections may be provided before and after one OFDM symbol. The time width of the guard interval is determined by the state of the propagation path, that is, the maximum delay time of the delayed wave that affects the demodulation. Then, the signal is converted into a serial signal, converted into a signal on the time axis while maintaining the orthogonality of each carrier on the frequency axis, and used as a transmission signal.

  The preamble / reference adding unit 24 adds a preamble signal or a reference signal to the head of a transmission signal such as an RTS, CTS, or DATA packet.

  A transmission signal for each antenna is converted into an analog baseband signal by each D / A converter 25 and further up-converted to an RF frequency band by each transmission analog processing unit 26, and then transmitted from each antenna 11. Transmitted to the MIMO channel.

  On the other hand, each reception system includes a reception analog processing unit 31 and an A / D converter 32 for each antenna, a synchronization acquisition unit 33, an FFT 34, a space separation unit 35, and a demodulation decoder 36.

  A signal received from each antenna 11 is down-converted from an RF frequency band to a baseband signal by each reception analog processing unit 31, and converted into a digital signal by each A / D converter 32.

  The digital baseband signal of each antenna system is collected by converting the received signal as serial data into parallel data in accordance with the synchronization timing detected by the synchronization acquisition unit 33 (including up to the guard interval here). Signals for one OFDM symbol are collected). Also, at this stage, timing error removal and frequency correction are performed on each digital baseband signal based on the frequency error estimation value.

  The FFT 35 converts the signal for the effective symbol length by Fourier transform into a signal on the time axis into a signal on the frequency axis, and decomposes the received signal into subcarrier signals.

The channel matrix acquisition unit 40 generates a channel matrix H for each subcarrier based on the FFT output of the preamble part of the packet, and calculates transmission / reception weights using this channel matrix. Specifically, since the reference signal corresponding to each MIMO channel is transmitted in a time division manner from the transmitter side (described later), the channel matrix acquisition unit 40 uses the transfer function acquired from each reference signal as each column vector. , And singular value decomposition of the channel matrix H into UDV ^H to give the reception weight U ^H to the space separation unit 35 and the transmission weight V to the spatial multiplexing unit 22.

  Further, in this embodiment, the channel matrix acquisition unit 40 performs the singular value rearrangement and transmission weight in the singular value decomposition in addition to the acquisition of the channel matrix H and the transmission / reception weight calculation processing by the singular value decomposition of the channel matrix H. Perform vector reordering. Thereby, the continuity of the frequency characteristics between the subcarriers arranged on the frequency axis can be maintained. However, details of the rearrangement process will be described later.

The space separation unit 35 uses the given reception weight U ^H to perform MIMO synthesis of the FFT output of the data portion of the packet for each subcarrier, and separates it into a plurality of independent MIMO channels.

  The demodulator / decoder 36 demodulates the original value from the modulation point on the phase space after the phase rotation correction.

  A multi-carrier transmission system represented by OFDM has a configuration in which a plurality of subcarriers are continuously arranged on the frequency axis. Therefore, interpolation is performed using the continuity of frequency characteristics in adjacent subcarriers. Communication quality can be improved.

However, in OFDM_MIMO communication in which OFDM modulation is applied to spatially multiplexed MIMO communication, transmission is performed on the transmitter side in order to perform transmission / reception weights for each subcarrier by singular value decomposition of the channel matrix H into UDV ^H. There is a problem that the continuity of the frequency characteristics between the subcarriers is lost when the data is multiplied by the transmission weight matrix V. The channel matrix H and the diagonal matrix D obtained by singular value decomposition are unique. On the other hand, there are an infinite number of combinations of the matrices U and V depending on the amount of phase rotation. This is because which V is used on the transmitter side is arbitrary, and continuity of frequency characteristics between subcarriers cannot be maintained (described above).

The transmission weight matrix V is composed of a combination of transmission weight vectors V ₁ , V ₂ ... For each MIMO channel to be spatially multiplexed (V = [V ₁ , V ₂ ,...)]. Similarly, the matrix U constituting the reception weight is composed of combinations of vectors U ₁ , U ₂ . Since the combination of U and V is determined depending on the setting of the reference phase of each vector, continuity between subcarriers is lost.

Therefore, in the present embodiment, the normalization of the phases of the transmission weight vectors V ₁ , V ₂ ... (Or U ₁ , U ₂ ...) Is performed to align the reference phases, so that the subcarriers arranged on the frequency axis are aligned. The continuity of frequency characteristics was compensated.

  The normalization of the phase of the transmission weight vector referred to here corresponds to aligning specific elements so as to have the same phase in each transmission weight vector. For example, when the elements of a vector are expressed in complex numbers, the phases of all the elements are rotated so that the same element of each vector is a positive constant.

In the case of a 2 × 2 MIMO communication system, the transmission weight matrix V consists of two 2 × 1 transmission weight vectors V ₁ and V ₂ . Here, when the transmission weight vector V ₁ = (a ₁ , a ₂ ) and mapping is performed on the complex plane as illustrated, each element a ₁ is placed so that the _first element a ₁ is placed on the real axis. and giving a uniform phase rotation amount θ in a _2, to phase rotation in a ₁ 'and a _2' (see Fig 2). This is carried out also for the other transmission weight vector V _2. This is a phase normalization process.

As another factor that hinders the continuity of the frequency characteristics between the subcarriers arranged on the frequency axis, the transmission weight vectors V ₁ , V ₂ ... Even if the transmission weight vectors V ₁ , V ₂ ... Are mistaken, the communication quality of the individual subcarriers is not greatly affected. However, the phase of the transmission weight vector changes due to the mistake, so that the continuity of the frequency characteristics cannot be expected. . Hereinafter, the cause of this mistake will be described.

The singular value decomposition of the channel matrix H obtains singular values λ ₁ , λ ₂ ... For each MIMO channel, rearranges these square roots into diagonal elements, generates a diagonal matrix D, and rearranges the diagonal elements. The process of rearranging the matrices U and V according to the above is included. In the case of a MIMO communication system to which multi-carrier modulation such as OFDM is applied, such processing is performed for each subcarrier.

The magnitudes of the singular values λ ₁ , λ ₂ ... Represent the communication quality of each MIMO channel. The process of singular value decomposition includes an operation of rearranging these singular values (in descending order). For example, the largest singular value obtained is sequentially assigned to the MIMO channel 1, the next largest singular value is assigned to the MIMO channel 2, and so on. In the multicarrier communication system, the singular values λ _{1, i} , λ _{2, i} ... (I indicates the i-th subcarrier) obtained for each subcarrier are arranged in a diagonal element in the order of value in a diagonal matrix. If the transmission weight vector is obtained by configuring D _i , the transmission weight vector should not be mixed up. However, when communication quality is antagonized between MIMO channels, that is, when the difference in singular values is small, the magnitude relationship is reversed due to the influence of noise, etc., and the transmission weight vector can be determined only by the magnitude of the singular values. The transmission weight vector may be confused. Even if the transmission weight vectors V ₁ , V ₂ ... Are mistaken, the communication quality of individual subcarriers is not greatly affected. However, the phase of the transmission weight vector changes due to the mistakes, so the continuity of the frequency characteristics between the subcarriers. Can not expect.

  Therefore, the inventors pay attention to the property that the transmission weight vector is highly correlated between adjacent subcarriers, and rearrange the next subcarrier to have a highly correlated transmission weight vector. By doing so, the mistake of transmission weight vectors due to noise is eliminated.

  FIG. 3 illustrates the internal configuration of the channel matrix acquisition unit 40 in this case. The channel matrix calculation unit 41 obtains the channel matrix H and HV (see Equation (8)) for each subcarrier using the reference signal included in the received packet. The reception weight acquisition unit 42 calculates a generalized inverse matrix of HV and gives this to the space separation unit 35 as a reception weight vector.

Further, the singular value decomposition unit 43 performs singular value decomposition on the channel matrix H (see Expression (6)). The process of singular value decomposition includes the operation of rearranging these singular values (in descending order). At the time of this rearrangement operation, the continuity of the frequency characteristics is hindered by the mistake of the transmission weight vectors V ₁ , V ₂ . There is a possibility that.

  Subsequently, rearrangement processing unit 44 performs a rearrangement operation on the transmission weight vector obtained for each subcarrier so as to maintain predetermined continuity between the subcarriers. Adjacent subcarriers have higher transmission weight vector correlation. By paying attention to this property and rearranging that the next subcarrier has a highly correlated transmission weight vector, it is possible to eliminate transmission weight vector mistakes due to noise. Rearrangement processing unit 44 obtains a transmission weight vector for a specific subcarrier in each MIMO channel. For other subcarriers, after performing singular value decomposition of the channel matrix independently, rearrangement of transmission weight vectors based on the correlation with transmission weight vectors determined in adjacent subcarriers on the frequency axis The operation is performed sequentially.

  Of course, if the transmission weight vector is rearranged, the processing load is high, so that the singular value for each subcarrier is monitored and the vector correlation calculation and the rearrangement process are performed only when the singular value approaches a certain threshold value or less. It may be.

  Here, in the OFDM_MIMO communication system, an advantage when the continuity of the frequency characteristics of the subcarriers arranged on the frequency axis is guaranteed will be described taking the case of obtaining the channel matrix H as an example.

  In a MIMO communication system, the basic operation is that a receiver acquires a channel matrix H using a reference signal sent from a transmitter and performs spatial separation using the inverse matrix as a receiving weight. In the SVD-MIMO communication system, for example, from the transmitter side, before the user data is spatially multiplexed and transmitted, a reference signal weighted with HV is time-division multiplexed to obtain a reception weight for each MIMO channel. Sent. On the receiver side, HV is acquired based on the reference signal, and its generalized inverse matrix is used for spatial separation.

  In SVD-MIMO to which OFDM modulation is applied as in IEEE 802.11n, it is necessary to perform a reference signal transmission and reception weight acquisition procedure for each subcarrier. FIG. 4 shows a configuration example of a data packet for transmitting a reference signal for each subcarrier in a 2 × 2 (that is, two transmitter and receiver antennas) MIMO system.

The illustrated packet has a configuration in which a HV-weighted reference signal is time-division multiplexed and transmitted after the preamble for synchronization acquisition, and then user data of each MIMO channel is spatially multiplexed and transmitted. ing. Since a communication system having a 2 × 2 antenna configuration is assumed here, the channel matrix H is a 2 × 2 matrix, and the transmission weights are two 2 × 1 transmission weight vector vectors V ₁ and consisting of V _2. Accordingly, two reference signals weighted by these transmission weight vectors V ₁ and V ₂ are transmitted in a time division manner.

On the receiver side, these weighted reference signals are received, HV is obtained, and a generalized inverse matrix of HV is used as a weight for reception, whereby one of the MIMO channels weighted by the transmission weight vector V ₁ is used. The user data and the user data on the other MIMO channel weighted with the transmission weight vector V ₂ can be spatially separated.

  In the case of the OFDM modulation scheme, such reference signal transmission and reception weight acquisition are performed for each subcarrier in the frequency axis direction. However, the packet shown in FIG. 4 has a configuration in which the area of the reference signal is expanded according to the number of MIMO channels. These reference signal portions become overhead, and if this area is large, the throughput is reduced.

  Therefore, a packet transmission method is conceivable in which the reference signal region is sent from the transmitter at intervals of one subcarrier, and the missing portion between them is interpolated on the receiver side to reduce the reference signal region. If the continuity of the frequency characteristics between the subcarriers arranged on the frequency axis is guaranteed, it is considered that highly accurate interpolation can be performed.

FIG. 5 shows a configuration example of a packet used in this case. Since the OFDM modulation scheme is applied, each subcarrier has a transmission weight vector. Here, the transmission weight vector in the j-th subcarrier of the i-th MIMO channel is represented as V _{(i, j)} .

  Weighting is performed on each subcarrier with a transmission weight vector of a reference signal in the frequency axis direction. According to the packet configuration shown in FIG. 5, the reference signal corresponding to MIMO channel 1 uses odd-numbered subcarriers, and the reference signal corresponding to the other MIMO channel 2 uses even-numbered subcarriers. In other words, even-numbered subcarrier reference signals are thinned out for MIMO channel 1, and odd-numbered subcarrier reference signals are thinned out for MIMO channel 2. Thus, since the area required for the reference signal is halved compared to FIG. 4, an improvement in throughput can be expected.

On the receiver side, in the case of the MIMO channel 1, the matrix HV related to the odd-numbered subcarriers, such as HV _(1,1) , HV _(1,3) , HV _(1,5) . The reception weight can be obtained directly from the inverse matrix. The HV _{(1, 2)} ... Relating to the even-numbered subcarriers of the MIMO channel 1 cannot be obtained directly from the reference signal, but the HVs adjacent to both sides of the HV _{(1, 2)} . It can be obtained by interpolation from _(1,1) and HV _(1,3) .

Similarly, with respect to the MIMO channel 2, the receiver side uses a matrix HV for even-numbered subcarriers such as HV _(2,2) , HV _(2,4) , HV _(2,6) . The reception weight can be obtained directly from the inverse matrix. In this MIMO channel 2, the HV _(2,3) ... Of the odd-numbered subcarriers is converted to HV _(2,2) and HV _{( 2,2) on} both sides by utilizing the continuity of the frequency characteristics between the subcarriers _. It can be obtained by interpolation from ₄₎ .

Now, consider a method of generating transmission weight vectors V ₁ and V ₂ for MIMO channel 1 and MIMO channel 2. The process of singular value decomposition includes an operation of rearranging these singular values (in descending order). That is, when the channel matrix H acquired based on the reference signal is subjected to singular value decomposition, the largest singular value corresponds to the MIMO channel 1, and the second largest singular value corresponds to the MIMO channel 2. The 2 × 1 column vector V corresponding to each singular value becomes the transmission weight vector V for the corresponding MIMO stream.

Such a singular value rearrangement operation is performed independently for each subcarrier. That is, the singular value decomposition is performed from the channel matrix H ₍₁₎ for the first subcarrier, and the V vector corresponding to the largest singular value is selected as the transmission weight vector for the MIMO stream 1. Next, singular value decomposition is performed from the matrix H ₍₃₎ for the third subcarrier, and the V vector for the largest singular value is selected as the transmission weight vector for MIMO stream 1.

There is no problem if the values of the first singular value and the second singular value are separated such as 3.0 and 1.0. However, when the values of singular values are close as in 2.1 and 1.9, this order may be switched due to the influence of noise. Since V vectors belonging to different singular values are orthogonal to each other, HV _{(1,2) is} interpolated from HV _(1,1) and HV _(1,3) on both sides while keeping this order in the wrong order. , The result is a completely strange value, and the receiving weight that starts from the bottom is naturally an inappropriate value. As a result, the reception characteristics are very deteriorated, and decent reception is impossible.

  In this embodiment, paying attention to the property that the correlation between transmission weight vectors is high between adjacent subcarriers, by performing a rearrangement that brings the transmission weight vector with high correlation to the next subcarrier, Such a mistake in transmission weight vectors is eliminated. Hereinafter, a processing procedure for correctly obtaining the reception weight based on the data packet in which the reference signal is thinned out at intervals of one subcarrier as shown in FIG. 5 will be described.

  After performing singular value decomposition independently for each subcarrier, one reference subcarrier is determined. For example, the subcarrier 1 is used as a reference.

  In subcarrier 1, the result of normal singular value decomposition is used as it is. That is, the V vector corresponding to the largest singular value is associated with MIMO channel 1 in the subcarrier, and the V vector corresponding to the second largest singular value is associated with MIMO channel 2 in the subcarrier.

Next, in subcarrier 3, singular value decomposition is performed independently of subcarrier 1, and vectors V _(1,3) and V _(2,3) are obtained. The order of the vectors given here is a tentative order. Correlation between V _(1,3) and V _(2,3) and transmission weight vector V _(1,1) associated with MIMO channel 1 in subcarrier 1 obtained earlier is obtained. Then, the one having a high correlation value is formally determined as the transmission weight vector V _(1,3) of the MIMO channel 1 in the subcarrier 3 concerned.

Further, in the subcarrier 5, singular value decomposition is performed independently of the above-described processing, and two 2 × 1 column vectors V _(1,5) and V _(2,5) are obtained. Is the order. Then, the correlation between V _(1,5) and V _(2,5) and the transmission weight vector V _(1,3) associated with the MIMO channel 1 in the subcarrier 3 obtained previously is obtained, respectively. _{Of (1,5)} and V _(2,5) , the one having a high correlation with V _(1,3) is formally determined as the transmission weight vector V _(1,5) of the MIMO channel 1 in the subcarrier 5. . Similarly, transmission weight vector V _(1,52) of MIMO channel 1 is obtained up to subcarrier 52 based on the correlation with the transmission weight vector in the latest subcarrier.

The other MIMO channel 2 is based on subcarrier 2. In subcarrier 2, the result of normal singular value decomposition is used as it is. That is, the vector V _(2,2) corresponding to the second largest singular value is made to correspond to the MIMO channel 2 in the subcarrier.

Next, the subcarrier 4 performs singular value decomposition independently of the subcarrier 2 to obtain vectors V _(1,4) and V _(2,4) . The order of the vectors given here is a tentative order. Correlation between V _(1,4) and V _(2,4) and transmission weight vector V _(2,2) associated with MIMO channel 1 in subcarrier 1 obtained earlier is obtained. Then, the one having a high correlation value is formally set as the transmission weight vector V _{(2, 4)} of the MIMO channel 2 in the subcarrier 4. Similarly, the transmission weight vector V _(2,52) of the MIMO channel 2 is obtained up to the subcarrier 52 based on the correlation with the transmission weight vector in the latest subcarrier.

  As described above, the MIMO channel 1 is rearranged based on the correlation value with the transmission weight vector in the latest subcarrier even when the V of the odd-numbered subcarrier is obtained and the order is difficult to determine with the value of the singular value. A correct transmission weight vector V can be obtained by sequentially performing the operations. Similarly, in the MIMO channel 2, V of even-numbered subcarriers is obtained, and a correct transmission weight vector V can be obtained by sequentially performing a rearrangement operation based on the correlation value.

  As can be seen from FIG. 5, the transmitter transmits the even-numbered subcarrier reference signal for the MIMO channel 1 and the odd-numbered subcarrier reference signal for the MIMO channel 2 for transmission. Is done. On the other hand, the receiver side can interpolate the wavy reception weight, which can be known from the thinned reference signal, using the continuity of the frequency characteristics of the subcarriers. That is, the HV of even-numbered subcarriers in MIMO channel 1 can be interpolated from the HV obtained by the odd-numbered subcarriers on both sides. On the other hand, although the even-numbered V is obtained in the MIMO channel 2, since the rearrangement of V is similarly performed using the correlation value, the odd-numbered HV of the MIMO channel 2 is obtained by interpolation on the receiver side. It becomes possible.

  Summarizing the above, instead of rearranging the singular value magnitude completely independently for each subcarrier, only a specific (reference) subcarrier is rearranged by the singular value magnitude, and the adjacent subcarrier is A method of rearranging using correlation values is employed in singular value decomposition. This can be realized even if the values of the singular values are very close to each other because the V vectors belonging to the singular values are orthogonal to each other, and thus the correlation is very low.

  Next, a transmission / reception procedure in the MIMO communication system in which the wireless communication apparatus shown in FIG. 1 is configured as a transmitter and a receiver will be described with reference to FIG. However, SVD_MIMO is applied to an OFDM modulation scheme in the 5 GHz band. Further, it is assumed that the number of antennas of the transmitter and the receiver is two, and two MIMO channels 1 and 2 are logically formed.

Step 1: A reference matrix is transmitted from a transmitter to a receiver, whereby a channel matrix H is obtained by the receiver.

Step 2: Transmit channel matrix H information from the receiver to the transmitter.

Step 3: The transmitter performs singular value decomposition on the received channel matrix H for each subcarrier. When there are 52 subcarriers, singular value decomposition is performed on the channel matrices H ₍₁₎ to H ₍₅₂₎ corresponding to the 52 subcarriers.

  When the transmitter transmits a packet, a reference signal weighted with a transmission weight vector of each MIMO channel is transmitted prior to the user data body in order to obtain a channel matrix H on the receiver side. At this time, as shown in FIG. 5, the reference signal corresponding to MIMO channel 1 uses odd-numbered subcarriers, and the reference signal corresponding to MIMO channel 2 uses even-numbered subcarriers.

V _{(1, j)} multiplied by the reference signal corresponding to MIMO channel 1 is based on subcarrier number 1. V _(1,1) is an eigenvector corresponding to the largest singular value. Then, with respect to the eigenvector subcarrier numbers 3 V _(1,3), of the V found by singular value decomposition independently of the subcarrier number 1 _(1,3) or V _(2,3), V _{( The} one having a high correlation with _1,1) is adopted as V _(1,3) . In the subsequent subcarriers, the same processing is repeated in which odd-numbered subcarriers having a high correlation with the V _{(1, j-2)} vector of the previous subcarrier are used as V _{(1, j).} (Where j is an odd number).

Also, V _{(2, j)} multiplied by the reference signal corresponding to MIMO channel 2 is based on subcarrier number 2. V _(2,2) is an eigenvector corresponding to the second largest singular value. The eigenvector V _(2,4) of subcarrier number 2 is correlated with V _(2,2) out of V _(1,4) or V _(2,4) obtained by independent singular value decomposition. The higher one is adopted as V _(2,4) . In the subsequent subcarriers, even-numbered subcarriers having a high correlation with the V _{(2, j-2)} vector of the previous subcarrier are adopted as V _{(2, j)} (where j is an even number). .

  Needless to say, the singular value for each subcarrier is monitored, and the vector correlation calculation and rearrangement processing may be performed only when the singular value approaches a certain threshold value or less.

Step 4: The weight vector determined in Step 3 is multiplied by the reference signal, and a packet as shown in FIG. 5 is transmitted from the transmitter to the receiver.

Step 5: The receiver acquires HV _(1,1) , HV _(1,3) ,..., HV _(1,51) corresponding to the MIMO channel 1 from the reference signal area. The even-numbered HV _{(1, j)} (where j is an even number) is obtained by interpolating from the adjacent ones. Further, HV _(2,2) , HV _(2,4) ,..., HV _(2,52) corresponding to the MIMO channel 2 are acquired. The odd-numbered HV _{(2, j)} (where j is an odd number) is obtained by interpolating from both adjacent ones.

Step 6: The receiver obtains a receiving weight by performing an inverse matrix operation from the obtained HV _{(i, j)} (see equation (8)). Then, the user data portion is spatially separated into MIMO channel 1 and MIMO channel 2 data to receive data.

  As already described, since the processing load is high when the transmission weight vectors are rearranged, the channel matrix acquisition unit 40 monitors the singular value for each subcarrier only when the singular value approaches a certain threshold value or less. Vector correlation calculation and rearrangement processing may be performed.

  In addition, when the channel matrix acquisition unit 40 is implemented by hardware, as one method for efficiently performing rearrangement based on correlation of transmission vectors, the correlation between adjacent subcarriers is obtained in parallel and adjacent to each other. It may be considered that 52 subcarriers are matched after the relationship between the subcarriers is determined first.

  FIG. 7 shows a configuration example of the rearrangement processing unit 44 in this case. In the figure, for simplification, the case where the number of subcarriers is eight is shown. First, the transmission vectors V are rearranged using correlation values in the relationship between subcarriers adjacent on the frequency axis. Subsequently, rearrangement is performed on the basis of the correlation between the rearranged ones. Thereby, parallelization of the rearrangement process can be realized, and the processing speed can be increased.

  The present invention has been described in detail above with reference to specific embodiments. However, it is obvious that those skilled in the art can make modifications and substitutions of the embodiment without departing from the gist of the present invention. That is, the present invention has been disclosed in the form of exemplification, and the contents described in the present specification should not be interpreted in a limited manner. In order to determine the gist of the present invention, the description of the scope of claims should be considered.

FIG. 1 is a diagram schematically showing a configuration of a wireless communication apparatus according to an embodiment of the present invention. FIG. 2 is a diagram for explaining the phase normalization processing of the transmission weight vector. FIG. 3 is a diagram illustrating an internal configuration of the channel matrix acquisition unit 40. FIG. 4 is a diagram illustrating a configuration example of a data packet for transmitting a reference signal for each subcarrier in a 2 × 2 MIMO system. FIG. 5 is a diagram showing a configuration example of a data packet when a reference signal region is transmitted at intervals of one subcarrier. FIG. 6 is a diagram showing a transmission / reception procedure in the MIMO communication system. FIG. 7 is a diagram for explaining a method of efficiently performing rearrangement based on correlation of transmission vectors. FIG. 8 is a diagram conceptually showing the MIMO communication system. FIG. 9 is a diagram conceptually showing the SVD-MIMO transmission system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 ... Antenna 12 ... Switch 21 ... Modulation encoding part 22 ... Spatial multiplexing part 23 ... IFFT
24 ... Preamble / reference adding unit 25 ... D / A converter 26 ... Analog processing unit for transmission 31 ... Analog processing unit for reception 32 ... A / D converter 33 ... Synchronization acquisition unit 34 ... FFT
35 ... Spatial separation unit 36 ... Decoder 40 ... Channel matrix acquisition unit 41 ... Channel matrix operation unit 42 ... Reception weight acquisition unit 43 ... Singular value decomposition unit 44 ... Rearrangement processing unit

Claims (14)

A wireless communication device that includes a plurality of antennas and performs OFDM (orthogonal frequency division multiplexing) transmission of spatially multiplexed signals on a plurality of spatial channels formed in pairs with a communication device having a plurality of antennas,
Channel matrix acquisition means for acquiring a channel matrix based on a received signal;
Singular value decomposition means for deriving a singular value decomposition of a channel matrix to obtain a transmission weight vector and a reception weight vector in each spatial channel;
Transmitting means for spatially multiplexing transmission data using the transmission weight vector and performing OFDM transmission;
Receiving means for spatially separating the OFDM signal received using the reception weight vector into signals for each spatial channel, and receiving processing;
The singular value decomposition means normalizes the phase of the transmission weight vector to align the reference phase of each element, and maintains the continuity of the frequency characteristics between the subcarriers arranged on the frequency axis. Reordering the transmission weight vector obtained for each
A wireless communication apparatus. A wireless communication device that includes a plurality of antennas and performs OFDM (orthogonal frequency division multiplexing) transmission of spatially multiplexed signals on a plurality of spatial channels formed in pairs with a communication device having a plurality of antennas,
Channel matrix acquisition means for acquiring a channel matrix based on a received signal;
Singular value decomposition means for deriving a singular value decomposition of a channel matrix to obtain a transmission weight vector and a reception weight vector in each spatial channel;
Transmitting means for spatially multiplexing transmission data using the transmission weight vector and performing OFDM transmission;
Receiving means for spatially separating the OFDM signal received using the reception weight vector into signals for each spatial channel, and receiving processing;
The singular value decomposition means obtains a transmission weight vector for a specific subcarrier in each spatial channel, performs singular value decomposition independently for other subcarriers, and then determines the adjacent subcarriers on the frequency axis. Transmission obtained for each subcarrier so as to maintain the continuity of the frequency characteristics between the subcarriers arranged on the frequency axis by sequentially rearranging the transmission weight vectors based on the correlation with the transmitted transmission weight vectors. Reordering the weight vectors,
A wireless communication apparatus. The singular value decomposition means monitors a singular value for each subcarrier, and performs correlation calculation and rearrangement processing of transmission weight vectors when the singular values of adjacent subcarriers approach a threshold value or less.
The wireless communication apparatus according to claim 2. The singular value decomposition means obtains the correlation between adjacent subcarriers in parallel, first determines the relationship between adjacent subcarriers, and then matches the entire subcarrier.
The wireless communication apparatus according to claim 1 , wherein the wireless communication apparatus is a wireless communication apparatus. The transmission means adds a reference signal weighted with a transmission weight vector for each spatial channel to a transmission packet in order to obtain a reception weight vector on the communication partner side.
The wireless communication apparatus according to claim 1 , wherein the wireless communication apparatus is a wireless communication apparatus. The transmission means weights the subcarriers of the reference signals arranged on the frequency axis by alternately using the transmission weight vector for each spatial channel in the subcarrier.
The wireless communication apparatus according to claim 5 . The reception means interpolates a reception weight vector for a spatial channel that could not be obtained in a certain subcarrier based on a reception weight vector for the spatial channel obtained in a subcarrier adjacent on the frequency axis.
The wireless communication apparatus according to claim 6 . A wireless communication method for performing OFDM (orthogonal frequency division multiplexing) transmission of spatially multiplexed signals on a plurality of spatial channels formed by a pair of communication devices having a plurality of antennas,
A reception step of spatially separating an OFDM signal received using the reception weight vector into signals for each spatial channel, and receiving processing;
A channel matrix obtaining step for obtaining a channel matrix based on a received signal;
A singular value decomposition step of determining a transmission weight vector and a reception weight vector in each spatial channel by singular value decomposition of the channel matrix;
Have
In the singular value decomposition step, the phase of the transmission weight vector is normalized to align the reference phase of each element, and the continuity of the frequency characteristics between the subcarriers arranged on the frequency axis is maintained. Reordering the transmission weight vector obtained for each
A wireless communication method. A wireless communication method for performing OFDM (orthogonal frequency division multiplexing) transmission of spatially multiplexed signals on a plurality of spatial channels formed by a pair of communication devices having a plurality of antennas,
A reception step of spatially separating an OFDM signal received using the reception weight vector into signals for each spatial channel, and receiving processing;
A channel matrix obtaining step for obtaining a channel matrix based on a received signal;
A singular value decomposition step of determining a transmission weight vector and a reception weight vector in each spatial channel by singular value decomposition of the channel matrix;
Have
In the singular value decomposition step, a transmission weight vector is obtained for a specific subcarrier in each spatial channel, and singular value decomposition is performed independently for other subcarriers, and then determined on adjacent subcarriers on the frequency axis. Transmission obtained for each subcarrier so as to maintain the continuity of the frequency characteristics between the subcarriers arranged on the frequency axis by sequentially rearranging the transmission weight vectors based on the correlation with the transmitted transmission weight vectors. Reordering the weight vectors,
A wireless communication method. In the singular value decomposition step, a singular value for each subcarrier is monitored, and when a singular value between adjacent subcarriers approaches a threshold value or less, a correlation calculation and rearrangement process of a transmission weight vector is performed.
The wireless communication method according to claim 9 . In the singular value decomposition step, the correlation between adjacent subcarriers is obtained in parallel, the relationship between adjacent subcarriers is determined first, and then the entire subcarrier is matched.
The wireless communication method according to claim 8, wherein the wireless communication method is a wireless communication method. A reference signal weighted with a transmission weight vector for each spatial channel in order to obtain a reception weight vector on the communication partner side, further comprising a transmission step of spatially multiplexing transmission data using the transmission weight vector and performing multicarrier transmission Is added to the transmitted packet,
The wireless communication method according to claim 8, wherein the wireless communication method is a wireless communication method. In the transmission step, each subcarrier of the reference signals arranged on the frequency axis is weighted by alternately using a transmission weight vector for each spatial channel in the subcarrier.
The wireless communication method according to claim 12 . In the reception step, the spatial channel reception weight vector that could not be obtained in a certain subcarrier is interpolated based on the spatial channel reception weight vector obtained in the adjacent subcarrier on the frequency axis.
The wireless communication method according to claim 13.

Images