JP4105175B2 - Wireless communication apparatus and method - Google Patents

Description

In a wireless system that performs communication via a wireless line, the present invention provides channel tracking (transmission) corresponding to a change in transfer function over time in order to suppress a decrease in reception quality due to a change in transfer function when the packet length to be transmitted is long. The present invention relates to a technique for performing function tuning.
In particular, the present invention uses the same frequency channel, transmits independent data from a plurality of different transmission antennas, receives signals using a plurality of reception antennas, and based on a transfer function matrix between the transmission and reception antennas. The present invention relates to a receiving technique for realizing good transmission characteristics while suppressing a circuit scale in a high-speed wireless access system that realizes wireless communication by demodulating data on a receiving station side.
In addition, the present invention is used particularly for increasing the transmission speed of a high-speed wireless access system using the 2.4 GHz band or the 5 GHz band.

In recent years, the IEEE802.11g standard, the IEEE802.11a standard, etc. have been remarkably spread as high-speed wireless access systems using the 2.4 GHz band or the 5 GHz band.
In these systems, a maximum transmission rate of 54 Mbps is realized. However, with the spread of wireless LAN, further increase in transmission rate is required.
For this purpose, MIMO (Multiple-Input Multiple-Output) technology is promising. This MIMO technology is such that different independent signals are transmitted on the same channel from a plurality of transmitting antennas on the transmitting station side, and signals are received using the same plurality of antennas on the receiving station side, between each transmitting antenna / receiving antenna. The transfer function matrix is obtained, the independent signal transmitted from each antenna on the transmitting station side is estimated using this matrix, and the data in the transmitted signal is reproduced.

Here, in the MIMO wireless transmission / reception system, a case is considered where N signals are transmitted using N transmitting antennas and the transmitted N signals are received using M antennas.
First, there are N × M transmission paths between the antennas of the transmitting station and the receiving station, and the transfer function when transmitted from the i-th transmitting antenna and received by the j-th receiving antenna is h _{j, i.} A matrix of M rows and N columns having this as the (j, i) th component is denoted as H. Further, let t _{i be} a transmission signal from the i-th transmitting antenna, Tx be a column vector whose components are (t ₁ , t ₂ , t ₃ ,..., T _N ), and r be a received signal at the j-th receiving antenna. _j is a column vector whose components are (r ₁ , r ₂ , r _3, ..., r _M ), and R _j is the thermal noise of the jth receiving antenna (n ₁ , n ₂ , n ₃ ,. ... A column vector whose component is n _M ) is expressed as n.
In this case, the following relationship (1) is established.
Rx = H × Tx + n (1)

Therefore, there is a need for a technique for estimating the transmission signal Tx based on the signal Rx received on the receiving station side. As the most basic one of the MIMO technology, there is a method generally called a ZF (Zero Forcing) method. (See Non-Patent Document 1)
Here, an inverse matrix H ⁻¹ of the transfer function matrix is obtained for (Equation 1) above, and a process of multiplying it from the left of both sides of the equation is performed. As a result, the following equation (2) is obtained.
H ⁻¹ × Rx = Tx + H ⁻¹ × n (2)
That is, when the signals received by the respective receiving antennas are combined and processing for removing interference caused by signals from other than the desired transmitting antenna is performed, a minute thermal noise H ⁻¹ × n is added to the actual transmitted signal vector Tx. A signal point is obtained.
Here, when a signal subjected to multilevel modulation such as BPSK, QPSK, 16QAM, or 64QAM is used as a transmission signal, possible signal points (signals obtained by mapping a digital signal by multilevel modulation) are discontinuous. It is.

Therefore, a hard decision process for searching for a signal point with the shortest Euclidean distance on the transmission constellation is performed for H ⁻¹ × Rx, and a true transmission signal is estimated.
In the above ZF method, when the thermal noise term H ⁻¹ × n is sufficiently small and the components for each transmitting antenna can be assumed to be uniform among the transmitting antennas, good signal reproduction characteristics are expected. it can.
However, in general, this assumption does not hold, and the expected value of the absolute value of the thermal noise H ⁻¹ × n for each reception antenna is different for a certain transfer function matrix (transmission characteristics between the transmission antennas are different).
Furthermore, if the transfer function matrix H is a matrix having no inverse matrix (or the determinant of the matrix is very small), the estimation of the transmission signal becomes very unstable.
In the situation described above, there is a possibility that the reception characteristics of the receiving antenna at the receiving station are significantly degraded.
As a method for solving such problems, a method having the most excellent characteristics is a method called an MLD method. (See Non-Patent Document 2)

In this MLD method, first, when the modulation method of a transmission signal from each transmitting antenna is determined, the number of signal points that can be taken by a signal transmitted from one antenna (hereinafter referred to as N _max ) is determined. There are N _max ^N types of variations of signal vectors transmitted across the N antennas.
Further, in the MLD method, prediction of received signals when the signals are transmitted is performed on all possible candidates (total of N _max ^N types) as transmission signals, and the most actual reception among them is performed. A signal point closest to the signal is selected as a signal point with the highest estimation accuracy. That is, if a transmission signal candidate Tx arbitrarily selected, for example, the kth transmission signal candidate is represented by Tx ^[k] , the Euclidean distance E defined by the following equation (3) is minimized. Select a value for k.
E = (Rx−H × Tx ^[k] ) ^H × (Rx−H × Tx ^[k] ) (3)
Note that ^{MH with} respect to the matrix M indicates a matrix that is Hermitian conjugate of the matrix M. With the above processing, the MLD method can perform stable reception processing for any matrix H, and reception characteristics are greatly improved compared to the ZF method.

Here, FIG. 2 shows a configuration of a transmission section of a transmission station to which the MIMO technique in the prior art is applied. In this figure, the transmission unit of the transmission station includes a data division circuit 101, a preamble assignment circuit 102-1, modulation circuits 103-1 to 103-N, radio units 104-1 to 104-N, and transmission antennas 105-1 to 105. -N.
As an example, a case where the transmitting station transmits N systems of data using N transmission antennas will be described as an example.

When data to be transmitted is input from the external circuit, the data dividing circuit 101 separates this data into N systems and outputs each to different preamble circuits.
For example, the data division circuit 101 outputs the first system data to the preamble provision circuit 102-1. As a result, the preamble circuit 102-1 is input to the modulation circuit 103-1 (Chl) with the preamble signal applied.
In the modulation circuit 103-1, predetermined modulation (multilevel modulation such as BPSK, QPSK, 16QAM, 64QAM, etc.) is performed on the data to which the input preamble signal is added, and the modulated transmission signal is wirelessly transmitted. Output to unit 104-1.

The radio unit 104-1 converts the transmission signal into a radio frequency and transmits the radio signal to the reception station via the transmission antenna 105-1 (radiated as a radio wave).
Similarly to the above-described “Ch1” system, the data of the second system output from the data dividing circuit 101 is converted into a preamble assigning circuit 102-2, a modulation circuit 103-2 (Ch2), a radio unit 104-2, And processed by the radio unit 104-2 and transmitted from the antenna 105-2.
Similarly, as in the above-described “Ch1” and “Ch2” systems, the N-th system data output from the data dividing circuit 101 is converted into a preamble adding circuit 102-N, a modulation circuit 103-N (ChN), and a radio unit. 104-N and the radio unit 104-N are processed and transmitted from the antenna 105-N.
Thereby, the data to be transmitted divided by the data dividing circuit 101 is individually transmitted from different antennas (105-1 to 105-N).

Next, FIG. 3 shows the configuration of the receiving unit of the receiving station in the prior art.
The reception unit of the reception station includes reception antennas 111-1 to 111 -M, radio units 112-1 to 112 -M, a channel estimation circuit 113, a reception signal management unit 114, a transfer function matrix management circuit 115, a signal detection circuit 116, The data composition circuit 117 is configured.
Further, the first reception antenna 111-11 to the Mth reception antenna 111-M individually receive the reception signals.

The channel estimation circuit 113 inputs the signal (packet) transmitted from the transmission unit via the radio units 112-1 to 112-M.
Then, the channel estimation circuit 113 determines the transfer function between each of the transmission antennas 105-1 to 105-N and the reception antennas 111-1 to 111-M based on the reception status of the predetermined preamble signal given on the transmission side in the packet. (Pilot data in the preamble signal is determined in advance by both the transmission and reception units, and a transfer function is obtained from the reception status of this pilot data).
The channel estimation circuit 113 outputs the acquired information h _{j, i} of each transfer function to the transfer function matrix management circuit 115 and outputs the data signal subsequent to the preamble signal to the reception signal management circuit 114 one symbol at a time. .
Here, the transfer function matrix management circuit 115 manages a matrix composed of h _{j, i} input from the channel estimation circuit 113 as a transfer function matrix H.

Then, the reception signal management circuit 114 receives the data signal input in symbol units using the reception signals (r ₁ , r ₂ ,..., R _M ) of the reception antennas 111-1 and 111 _-M as components. The signal vector Rx is once managed by an internal storage unit.

The signal detection circuit 116 estimates the most likely transmission signal from the relationship between the transfer function matrix managed by the transfer function matrix management circuit 115 and the reception signal managed by the reception signal management circuit 114. As the method used at this time, other methods including the methods defined in Non-Patent Document 1, Non-Patent Document 2, and the like may be used. Here, it will be generalized and will be described as the signal detection circuit 116.
The data as the estimated transmission signal output from the signal detection circuit 116 is processed in a time series continuously over a plurality of symbols. However, after receiving a series of data, the data synthesis circuit 117 synthesizes symbols to generate data. Is reconstructed and output as
S. Kurosaki et. Al., “A SDM-COFDM Scheme Employing a Simple Feed-Forward Inter-Channel Interference Canceller for MIMO Based Broadband Wireless LANs”, IEICE TRANS. COMMUN ・, Vol.E86 B. 2003 A.van Zelst et.al., “Space Division Multiplexing (SDM) for OFDM Systems”, Proc, VTC2000 Spring, Vol. 2, pp.1070 -1074

  In the above reception processing, when the data length of the input packet is long, the transfer function matrix H extracted by the preamble at the beginning of the packet changes with time, and the symbol located at the back of the packet is reproduced with high accuracy. May not be possible. In the prior art, information on the transfer function used for signal detection is acquired at the beginning of the packet and the value is assumed to be almost unchanged in the packet. However, in order to improve transmission efficiency in the MAC layer, the packet length is This assumption is broken when expanding and sending long packets. In particular, IEEE802.11n, which has been standardized to improve the transmission speed by expanding IEEE802.11a, 802.11g, etc., which are high-speed wireless access systems, is actually being studied in this way. In order to be able to receive a long packet stably even under difficult operating conditions, there was a need.

An example of a countermeasure for this is the method described in “Asai et al.,“ Simplified Blind Path Tracking Method for MIMO-OFDM ”, IEICE Technical Report, RCS2004-290”. The principle is briefly shown below. The transfer function information is obtained by comparing the known transmission signal with the actual reception state of the signal. For the data portion other than the preamble, there is no reason to know the true transmission signal on the reception side, but it is possible to estimate it in the process of signal detection processing at reception.
For example, the case where the number of transmission / reception antennas is 3 and the number of MIMO multiplexing is also 3 is described as an example. A column vector of 3 rows and 1 column having signals transmitted from three transmitting antennas as components at a certain time (symbol) is defined as T ₁ . In addition, a column vector whose components are reception signals at three reception antennas for this signal is R ₁ . Similarly, it is assumed that T ₂ and R ₂ , and T ₃ and R ₃ are obtained as combinations of transmission signals and reception signals at different times. Assume that the transfer function matrix (3 rows and 3 columns) at this time is H. A replica matrix of a 3 × 3 transmission signal in which transmission signal vectors T ₁ , T ₂ , and T ₃ are arranged in each column is represented as T _rep, and further, received signal vectors R ₁ , R ₂ , and R ₃ are respectively represented as T _rep. When a matrix of received signals of 3 rows and 3 columns arranged in columns is expressed as R _rep , the following equation (4) is given.
R _rep = H · T _rep + N (4)
Here, the last term is a matrix related to thermal noise. Therefore, if this term is ignored, the transfer function matrix can be estimated by the following equation (6).
H ≒ R _rep · T _rep ^-1 (5)
Therefore, it is possible to estimate the transfer function matrix at that moment from the multiplication of the reception signal matrix configured based on the reception signal and the inverse matrix of the replica matrix of the transmission signal.

In this case, in FIG. 3, the transfer function matrix management circuit 115 needs to calibrate the transfer function at predetermined intervals, and re-encodes and re-modulates the decoded information signal sequence. Thus, a replica matrix of the transmission signal is generated.
Further, the transfer function matrix management circuit 115 reads the reception signal matrix at the time corresponding to this replica matrix from the reception signal matrix group stored in the reception signal management circuit 14.
Then, the transfer function matrix management circuit 115 obtains a transfer function estimation matrix by multiplying the inverse matrix of the replica matrix from the right of the received signal matrix.
In this way, the transfer function matrix management circuit 115 replaces the transfer function matrix H stored at that time every predetermined time with the obtained transfer function estimation matrix so that the transfer function matrix (that is, the channel) Tracking can be performed.

Since the above-described channel tracking method is MIMO, it is necessary to improve estimation accuracy. A replica signal of a transmission signal is generated for information of a plurality of symbols, and a transfer function is generated using the inverse matrix and the reception signal matrix. The estimation matrix is estimated.
However, in this channel tracking method, the estimation accuracy depends on the characteristics of the replica matrix to be used, a plurality of symbols are required to obtain the replica matrix, and the update period is also required to be more than a plurality of symbol periods. There is a drawback that it cannot support tracking. Furthermore, when a transmission signal vector at a certain time is represented by a linear combination of one or a plurality of transmission signal vectors at other times, the replica matrix of the transmission signal does not have an inverse matrix. In this case, the transfer function matrix cannot be obtained, and the update process cannot be performed.

In addition, the channel tracking method requires 4 × 4 inverse matrix calculation when there are a large number of multiplexed MIMOs, for example, if N transmission antennas, for example, 4 antennas are used and quadruplex multiplexing is performed. As a result, the scale of the arithmetic circuit for performing the calculation becomes large, which has been an obstacle to mounting all functions in one chip. Furthermore, a delay associated with the arithmetic processing occurs, and more time is required for channel tracking.
Therefore, an object of the present invention is to perform channel tracking with high speed and high estimation accuracy in a short symbol period compared to the conventional example when performing wireless communication using the MIMO technology, with a realistic circuit scale and calculation amount. It is to provide a wireless communication apparatus that can be realized by the above.

In order to solve the above problem, the wireless communication system of the present invention includes N (N ≧ 2, N is an integer) transmission antennas that multiplex and transmit a plurality of signal sequences on the same frequency channel. And a receiving station having M (M ≧ 1, M is an integer) receiving antennas that receive a transmitted radio signal and perform reception processing by separating the signal into a plurality of signal sequences. A wireless communication system capable of (Multiple Input Multiple Output) communication, wherein the transmitting station includes user data dividing means for dividing user data into N systems,
A signal sequence generating means for generating a signal sequence of N systems by giving a signal of an individual known pattern to the data divided into the N systems, and the signal simultaneously at the same frequency using N transmission antennas Signal transmitting means for superimposing and transmitting the sequence, and the receiving station individually receives radio signals using the M receiving antennas, and a known pattern given to the received signal in the previous period. Transfer function obtaining means for obtaining M × N sets of transfer functions h _{j, i} between the i-th antenna of the transmitting antennas and the j-th antenna of the receiving antennas using the signal of A matrix of M rows and N columns, that is, a transfer function matrix having the transfer function h _{j, i} as the (j, i) component, that is, a transmission function matrix N, a transmission signal sequence vector T of N rows and 1 column, Single symbol received by M antennas When the received signal is expressed as a received signal sequence vector R of M rows and 1 column, the transmission function sequence H and the transmitted signal sequence vector transmitted by the transmitting station using the received signal sequence vector R for each symbol. Transmission signal estimation means for obtaining an estimated transmission signal sequence vector T ′, which is an estimated value of T, and a matrix of N rows and N columns, and a predetermined k-th (1 ≦ k ≦ N: k is an integer) column of the matrix Test matrix storage means for storing a test matrix U in which the values of the eye components are all constants C (C is an arbitrary real or complex number) and an inverse matrix U ⁻¹ of the matrix;
Diagonal matrix generating means for generating a square diagonal matrix D of N rows and N columns in which each of N components of the estimated transmission signal sequence vector T ′ is a diagonal component and all non-diagonal components are zero; Inverse matrix obtaining means for obtaining an inverse matrix D ⁻¹ of a square diagonal matrix D; a product of the transfer function matrix H, the square diagonal matrix D and the test matrix U, that is, a matrix H × D · U of M rows and N columns A first matrix multiplying unit for obtaining a converted received signal sequence vector R ′ for multiplying each component of the received signal sequence vector R by the constant C, and converting the kth row of the matrix of M rows and N columns Replica matrix generation means for generating a replica matrix H _rep replaced by the received signal column vector R ′, a product of the replica matrix H _rep , the inverse matrix U ^{−1 of the} test matrix, and the square diagonal matrix D ⁻¹ , that is, M second row to obtain the matrix _H rep ^{· U -1 · D} -1 rows and N columns Multiplication means, new transfer function matrix generation means for generating a new transfer function matrix as a linear combination of the matrix and the transfer function matrix, and the transfer function matrix used for the previous reception processing is updated to the new transfer function matrix And a transfer function updating means.
As a result, when receiving a packet having a long packet length in a MIMO communication apparatus, a tracking process (channel tracking) of fluctuation due to the passage of time of the transfer function estimated by the preamble is performed at the timing when one symbol is input. Compared to the processing using multiple symbols as in the conventional tracking method, channel tracking can be performed in a short period, so that the transfer function variation factor between multiple symbols can be reduced. It is possible to estimate a transfer function with higher accuracy without including it.

  In the wireless communication apparatus of the present invention, the test matrix U is characterized in that the column vectors constituting the matrix are orthogonal to each other.

Alternatively, in the wireless communication apparatus of the present invention, it is preferable that the test matrix U is a unitary matrix or a matrix obtained by multiplying each component of the unitary matrix by a constant.
In the present invention, an inverse matrix of the test matrix U is used, and this inverse matrix is multiplied by the inverse of the determinant of the matrix U as a coefficient.
By applying the above condition to the matrix U, it is possible to suppress this coefficient to a small value, and as a result, it is possible to perform channel estimation while suppressing unnecessary noise amplification.

In the wireless communication apparatus of the present invention, the new transfer function matrix generation means uses a forgetting factor μ (0 <μ ≦ 1: μ is a real number), and (1-μ) times the transfer function H and a matrix H _A new transfer function matrix is generated by linearly synthesizing μ times _rep · U ⁻¹ · D ⁻¹ .
As a result, the wireless communication apparatus of the present invention suppresses a sudden change in the transfer function, that is, the newly estimated transfer function matrix H _rep · U ⁻¹ · D ⁻¹ has low estimation accuracy due to the influence of noise or the like. If the transfer function is significantly different from the actual transfer function, it is possible to adjust the transfer function with the communication quality to the extent that error correction can be performed, that is, to perform channel tracking, preventing drastic deterioration of communication characteristics due to complete replacement. .

The radio communication apparatus of the present invention is characterized in that the forgetting factor μ is dynamically changed in proportion to an absolute value of the estimated transmission signal sequence vector T ′.
Thereby, in the wireless communication device of the present invention, when the absolute value of the estimated transmission signal sequence vector to be estimated is large, since the received signal strength is high, the estimated transfer function is closer to the actual transfer function, That is, since the estimation accuracy may be high, the forgetting factor μ is increased. On the other hand, when the absolute value of the estimated estimated transmission signal sequence vector is small, the received signal strength is low, so that it is strongly influenced by noise. Since the estimated transfer function is different from the actual transfer function, that is, it is considered that the estimation accuracy is low, the channel tracking accuracy can be improved by reducing the forgetting factor μ.

In addition, the receiving apparatus of the present invention receives a MIMO (N MIMO (N ≧ 2, N is an integer)) transmission station having N (N ≧ 2, N is an integer) transmitting a plurality of signal sequences on the same frequency channel. A receiving apparatus having M (M ≧ 1, M is an integer) receiving antennas that receive a radio signal transmitted by a multiple input (Multiple Input Multiple Output) communication method and perform reception processing by separating the signal into a plurality of signal sequences. A signal receiving means for individually receiving radio signals using the M receiving antennas, a signal having a known pattern given to the received signal as a reference signal, and the i-th antenna among the transmitting antennas; Transfer function acquiring means for acquiring M × N sets of transfer functions h _{j, i} between the receiving antenna and the j th antenna, and the transfer function h _{j, i} as the (j, i) component M-row N-column matrix, ie transfer function matrix When a transmission signal of an N system signal sequence is expressed as a transmission signal sequence vector T of N rows and 1 column, and a reception signal in units of symbols received by M antennas is expressed as a reception signal sequence vector R of M rows and 1 column, Transmission signal estimation means for obtaining an estimated transmission signal sequence vector T ′ which is an estimated value of the transmission signal sequence vector T transmitted by the transmission station using the transfer function matrix H and the received signal sequence vector R for each symbol. And the values of the components in the predetermined k-th column (1 ≦ k ≦ N: k is an integer) of the matrix are all constants C (C is an arbitrary real number or complex number). Test matrix storage means for storing a test matrix U and an inverse matrix U ⁻¹ of the matrix, and N components of the estimated transmission signal sequence vector T ′ as diagonal components and all off-diagonal components are zero Diagonal matrix generating means for generating a square diagonal matrix D of N rows and N columns The inverse matrix acquisition means for obtaining the inverse matrix D ^-1 of the positive side diagonal matrix D, the matrix H · D of the product i.e. M rows and N columns of the transfer function matrix H and the square diagonal matrix D the test matrix U First matrix multiplication means for obtaining U, and a converted received signal sequence vector R ′ that multiplies each component of the received signal sequence vector R by the constant C, and the k-th row of the matrix of M rows and N columns is generated. Replica matrix generation means for generating a replica matrix H _rep replaced with the converted received signal sequence vector R ′, a product of the replica matrix H _rep , the inverse matrix U ^{−1 of the} test matrix, and the square diagonal matrix D ⁻¹ That is, the second matrix multiplication means for obtaining the matrix H _rep · U ^-1 · D ^-1 of M rows and N columns, and a new transfer function matrix generation for generating a new transfer function matrix as a linear combination of the matrix and the transfer function matrix And used for previous reception processing The reach function matrix is characterized in that a transfer function updating means for updating to the new transfer function matrix.

In addition, the reception method of the present invention provides a MIMO (N MIMO (N ≧ 2, N is an integer)) transmission station equipped with MIMO (N ≧ 2, N is an integer) transmitting a plurality of signal sequences on the same frequency channel. A reception method for receiving wireless signals transmitted by a multiple input (Multiple Input Multiple Output) communication method using M reception antennas (M ≧ 1, M is an integer) that performs reception processing by separating the signal into a plurality of signal sequences. A signal reception process for individually receiving radio signals using the M reception antennas, and a signal having a known pattern added to the reception signal as a reference signal, and the i-th antenna of the transmission antenna and the reception Transfer function acquisition process for acquiring M × N sets of transfer functions h _{j, i} between the antenna and the j-th antenna, and M rows with the transfer function h _{j, i} as the (j, i) component An N-column matrix, that is, a transfer function matrix is represented as H When a transmission signal of an N system signal sequence is expressed as a transmission signal sequence vector T of N rows and 1 column, and a reception signal in units of symbols received by M antennas is expressed as a reception signal sequence vector R of M rows and 1 column, A transmission signal estimation process for obtaining an estimated transmission signal sequence vector T ′ that is an estimated value of the transmission signal sequence vector T transmitted by the transmission station using the transfer function matrix H and the received signal sequence vector R for each symbol. And the values of the components in the predetermined k-th column (1 ≦ k ≦ N: k is an integer) of the matrix are all constants C (C is an arbitrary real number or complex number). A test matrix storage process for storing the test matrix U and the inverse matrix U ⁻¹ of the matrix in the test matrix storage means, and N components of the estimated transmission signal sequence vector T ′ as diagonal components and non-diagonal components Generate an N-by-N square diagonal matrix D with all zeros A diagonal matrix generation process, positive side reverse matrix obtaining means for obtaining the inverse matrix D ^-1 of the diagonal matrix D, the transfer function matrix H the product i.e. M rows N of the square diagonal matrix D and the test matrix U A first matrix multiplication process for obtaining a matrix H, D, U of columns, and a converted received signal column vector R ′ for multiplying each component of the received signal column vector R by the constant C, a replica matrix generation process for the k-th row of the matrix generates a replica matrix H _rep was replaced with the converted received signal column vector R ', the square diagonal and inverse matrix U ^-1 of the replica matrix H _rep and the test matrix A second matrix multiplication process for obtaining a product of the matrix D- ¹ , that is, a matrix H _rep · U ^-1 · D ^-1 having M rows and N columns, and generating a new transfer function matrix as a linear combination of the matrix and the transfer function matrix New transfer function matrix generation process and The transfer function matrix which has been used in the process and having a transfer function updating process of updating to the new transfer function matrix.

Note that it is also preferable to apply the above method to a radio communication system using an orthogonal frequency division multiplexing (OFDM) modulation scheme using a plurality of subcarriers.
In particular, when MIMO processing and OFDM technology are applied and reception processing is performed on the receiving side using the MLD method or a method based thereon, it is necessary to perform a large number of matrix calculations when performing channel tracking processing. By performing this approximation process with a simple circuit having a small circuit scale, the circuit scale can be reduced and the power consumption can be reduced.

As described above in detail, according to the present invention, when performing highly efficient wireless communication using the MIMO technology, the channel tracking process is performed with one symbol, so that it is compared with the conventional channel tracking system. An effect of greatly reducing the circuit scale and the amount of calculation can be obtained.
As a result, according to the present invention, the circuit scale can be greatly reduced, so that the receiving circuit can be mounted in a one-chip LSI. Further, the above-described reduction in circuit scale and reduction in the amount of calculation can also be expected to have a secondary effect of directly reducing power consumption.

  The wireless communication system of the present invention estimates a transmission signal sequence vector T from a reception signal sequence vector R of one symbol in tracking of a transfer function in MIMO communication, and obtains a reception signal matrix corresponding to the estimated transmission signal sequence vector T. A part of the generated reception signal matrix is replaced with the reception signal sequence vector R to generate a reception signal replica matrix, and the reception signal replica matrix is inverted to the transmission signal matrix composed of the transmission signal sequence vector. Tracking processing is performed by multiplying the matrix and estimating the transfer function at the time when the received signal sequence vector R is received.

Hereinafter, a wireless communication apparatus according to an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the configuration of the embodiment.
Hereinafter, embodiments of the present invention will be described with reference to the drawings. The difference between the present invention and the prior art is in the configuration of the receiving unit, and the configuration on the transmission side of the present invention is the same as the conventional example already described.
Therefore, only the receiving station will be described below. Similar to the conventional method, a case where the transmitting station transmits N systems of data using N transmitting antennas is used as an example.

FIG. 1 is a diagram illustrating a configuration of a receiving unit of a receiving station in the embodiment of the present invention. In the figure, the receiving unit of the present invention includes receiving antennas 111-1 to 111 -M, radio units 112-1 to 112 -M, channel estimation circuit 113, received signal management circuit 114, transfer function matrix management circuit 115, signal detection. The circuit 116 and the data synthesis circuit 117 are provided as in the conventional example.
The difference between the present invention shown in FIG. 1 and the conventional example shown in FIG. 3 is that a diagonal matrix generation circuit 1, a diagonal inverse matrix generation circuit 2, a first matrix multiplication circuit 3, a test matrix storage circuit 4, a received signal conversion circuit 5, a replica matrix generation circuit 6, a second matrix generation circuit 7, a new transfer function matrix generation circuit 8, and a transfer function matrix update control circuit 9 are newly provided. Hereinafter, the newly provided circuit will be described.

The first reception antenna 111-1, the second reception antenna 111-2,..., And the Mth reception antenna 111-M individually receive reception signals.
The radio units 112-1 to 112 -M receive the received signals individually received by the first receiving antenna 111-1, the second receiving antenna 111-2, and the M-th receiving antenna 111 -M, respectively. Then, signal processing such as amplification and noise removal is performed and output to the channel estimation circuit 113.
The channel estimation circuit 113 uses, as a reference signal, a preamble signal composed of known pattern data given on the transmission side, and, as described above, N transmission antennas i (1 ≦ i ≦). N) and M × N sets of transfer functions (h _{j, i} ) between the receiving antennas j (1 ≦ j ≦ M) are acquired and output to the transfer function matrix management circuit 115.
The transfer function matrix management circuit 115 manages an input transfer function (h _{j, i} ) as a transfer function matrix H, which is a matrix of M rows and N columns, using matrix components.

The received signal management circuit 114 inputs a data signal subsequent to the preamble signal one symbol at a time from the channel estimation circuit 113 as a received signal vector Rx having received signals (r1, r2,..., RM) of the respective antennas as components. Once managed.
In the signal detection circuit 116, the estimated transmission signal Tx (N lines) that seems to be the most probable from the relationship between the transfer function matrix managed by the transfer function matrix management circuit 115 and the reception signal vector Rx managed by the reception signal management circuit 114. N-row and 1-column estimated transmission signal sequence vectors) transmitted by the antennas of (2) are obtained and output to the diagonal matrix generation circuit 1. As the transmission signal estimation method at this time, other methods may be used including the methods defined in Non-Patent Document 1, Non-Patent Document 2, and the like as described in the related art.

The test matrix storage circuit 4 has N rows and N columns and all the values of the components in a predetermined k-th column (1 ≦ k ≦ N: k is an integer) of the matrix are constants C (C is an arbitrary real number or complex number). A test matrix U and an inverse matrix U ⁻¹ of the test matrix U are stored.
The diagonal matrix generation circuit 1 generates an N-row N-column square diagonal matrix D in which all the non-diagonal components are “0” using the N components of the estimated transmission signal sequence vector Tx as diagonal components. Generate.
Here, when Tx {t1, t2,..., TN}, the diagonal components of the square diagonal matrix D are {d11, d22,..., DNN} = {t1, t2,.
The diagonal inverse matrix generation circuit 2 generates an inverse matrix D ⁻¹ of the square diagonal matrix D by a predetermined calculation. This inverse matrix can be easily obtained by replacing each diagonal component with the reciprocal of its value.

The first matrix multiplication circuit 3 calculates the product of the transfer function matrix H, the square diagonal matrix D, and the test matrix U, that is, calculates an M × N matrix H · D · U.
Here, by multiplying the square diagonal matrix D and the test matrix U, each diagonal component {t1, t2,..., TN} of the square matrix D is multiplied by the column component of the test matrix U. , Tx1, Tx2, Tx3,..., TxN corresponding to the column components of the test matrix U are obtained as virtual transmission signal matrices. Here, the first column of the test matrix U is a constant 1. For this reason, the first column of the virtual transmission signal matrix is a transmission signal vector estimated by the signal detection circuit 116. The other column vectors of the virtual transmission signal matrix are signals obtained by multiplying the transmission signal vector by each element of the test matrix.
In this way, when the product of the square diagonal matrix D and the test matrix U is regarded as a virtual transmission signal matrix, the matrix H · D · U becomes a virtual reception signal matrix for the virtual transmission signal.

The received signal conversion circuit 5 multiplies each component of the received signal sequence vector Rx by a constant C to generate a converted received signal vector Rx ′ in which each component is multiplied by C. In this example, all the first columns of the test matrix are 1, and the constant C is 1. Therefore, the C-fold process is not a process for converting the original value. In such a case, the reception signal conversion circuit 5 can be omitted.
The replica matrix generation circuit 6 performs processing to replace the k-th row (1 ≦ k ≦ M, k is an integer) of the virtual received signal matrix of M rows and N columns with the converted received signal sequence vector Rx ′, and converts the received signal Each component of the column vector Rx ′ generates a replica matrix H _{rep in} which each column component is a component. In this example, all the first columns of the test matrix are 1, and the process of replacing with the converted received signal sequence vector Rx ′ is performed on the first column of the virtual received signal matrix.
The second matrix multiplication circuit 7 reads the inverse matrix U ⁻¹ of the test matrix U from the test matrix storage circuit 4, and the inverse of the replica matrix H _rep , the read inverse matrix U ^−1, and the square diagonal matrix D. by multiplying the matrix ^{D -1,} determine the matrix _H rep ^{· U -1 · D} -1 of M rows and N columns.
That is, the second matrix multiplication circuit 7 multiplies the replica matrix H _rep of the received signal by the inverse matrix of the virtual transmission signal matrix to calculate a new estimated transfer function matrix Hest.

The new transfer function matrix generation circuit 8 reads the transfer function matrix Hnow currently stored in the transfer function matrix management circuit 115, and uses the predetermined forgetting coefficient μ (0 <μ ≦ 1), and the following equation (6) Thus, the transfer function matrix Hnew to be newly stored in the transfer function matrix management circuit 115 is calculated.
Hnew = (1-μ) · Hnow + μ · Hest (6)
The estimated transfer function matrix Hest multiplied by a predetermined forgetting factor μ and the transfer function matrix Hnow multiplied by “1-μ” are linearly combined by the above equation (4) to estimate the estimated transfer function Instead of using the matrix Hest as it is, the matrix Hest is not replaced, and the matrix Hest is adjusted gently while leaving the previous component.
In the transfer function matrix management circuit 115, the transfer function matrix update control circuit 9 overwrites the obtained transfer function matrix Hnew on the transfer function matrix Hnow and sets it as a new transfer function matrix Hnow. In addition, when the obtained transfer function matrix Hnew is insufficient as the estimation accuracy, the replacement can be controlled as necessary, such as deferring the replacement of the transfer function matrix Hnow.

Next, with reference to FIG.1 and FIG.2, operation | movement of the radio | wireless communications system which has a transmission part and a receiving part in embodiment mentioned above is demonstrated.
Here, in order to simplify the description, for example, four transmission antennas and four reception antennas are used, that is, the transmission unit includes transmission antennas 105-1 to 105-4, and the reception unit includes reception antenna 111-. 1 to 111-4, and other configurations are also configured corresponding to the number of antennas.
The transmission unit individually transmits data from the transmission antennas 105-1 to 105-4.

And the 1st receiving antenna 111-1-the 4th receiving antenna 111-4 each receive a received signal individually.
As a result, the radio units 112-1 to 112-4 receive the received signals Rx {r1, r2, r3, r4} respectively received by the first receiving antenna 111-1 to the fourth receiving antenna 111-4. The signal is input, subjected to signal processing such as amplification and noise removal, and output to the channel estimation circuit 113.
Next, the channel estimation circuit 113 uses a preamble signal composed of known pattern data given on the transmission side as a reference signal, and from the reception status of this reference signal, as described above, each of the four transmission antennas i ( 4 × 4 sets of transfer functions (h _{j, i} ) between 1 ≦ i ≦ 4) and each of the four receiving antennas j (1 ≦ j ≦ 4) are acquired and output to the transfer function matrix management circuit 115 To manage.

The received signal management circuit 114 receives a data signal subsequent to the preamble signal one symbol at a time from the channel estimation circuit 113, and includes a received signal (r1, r2, r3, r4) of each antenna as a component. Store the vector Rx.
Next, the signal detection circuit 116 reads the reception signal vector Rx from the reception signal management circuit 114 and reads the transfer function matrix Hnow from the transfer function matrix management circuit 115 to obtain an estimated value of the transmission signal, 4 rows and 1 column. The estimated transmission signal sequence vector Tx {t1, t2, t3, t4} is output.

Next, the diagonal matrix generation circuit 1 uses four components of the virtual transmission signal sequence vector Tx as diagonal components, and is a 4 × 4 square diagonal matrix in which all the non-diagonal components are “0”. D (matrix D in equation (7) shown below) is generated.
Next, the first matrix multiplication circuit 3 reads the test matrix U from the test matrix storage circuit 4, and, as shown in the following equation (7), the test matrix U ((7 ) To multiply the matrix U) in the equation from the right side to obtain the virtual transmission signal matrix Tr.

Here, as shown in the equation (7), the virtual transmission signal matrix Tr {Tx1, Tx2, Tx3, Tx4} is the first column of the virtual transmission signal vector Tx1 = {t1, t2, t3, t4}, Tx2. = {T1, -t2, -t3, t4}, Tx3 = {t1, t2, -t3, -t4}, Tx4 = {t1, -t2, t3, -t4}.
Then, the first matrix multiplication circuit 3 multiplies Hnow read from the transfer function matrix management circuit 115 from the left side of the virtual transmission signal matrix Tr, that is, a 4 × 4 matrix Hnow · D · U, that is, A virtual received signal matrix Rr is calculated.

Next, the reception conversion circuit 5 multiplies each component of the reception signal sequence vector Rx input from the reception signal management circuit 114 by a constant “C” to obtain a conversion reception signal vector Rx ′ {C · r 1, C · r 2. , C · r3, C · r4}. Here, since all the first columns of the test matrix U are one, C = 1. That is, in this case, R′x = Rx, and the processing of the reception switching circuit 5 can be omitted.
Then, the replica matrix generation circuit 6 performs processing to replace the first column of the received signal matrix Rr of 4 rows and 4 columns with the converted received signal sequence vector Rx ′ as shown in the following equation (8), A matrix H _rep is generated.

Then, the diagonal inverse matrix circuit 2 calculates the ^{(D -1} in shown later ⁽⁹⁾⁾ diagonal matrix inverse matrix D of a square diagonal matrix D, which is calculated in the generator 1 ^-1.
Then, the second matrix calculation circuit 7 reads the inverse matrix U ⁻¹ of the test matrix U from the test matrix storage circuit 4 and converts the inverse matrix U ⁻¹ to a square diagonal matrix as shown in the following equation (9). By multiplying from the left side of the inverse matrix D ⁻¹ of D, the inverse matrix Tr ⁻¹ (= U ⁻¹ · D ⁻¹ ) of the virtual transmission signal matrix Tr is calculated.

Next, the second matrix multiplication circuit 7 multiplies the inverse matrix Tr ⁻¹ by a 4 × 4 matrix H _rep from the left to obtain an estimated transfer function matrix Hest as a 4 × 4 matrix H _rep. -U- ¹ * D- ¹ is calculated ^| required.
Then, the new transfer function matrix generation circuit 8 reads the transfer function matrix Hnow currently stored in the transfer function matrix management circuit 115 and newly stores it in the transfer function matrix management circuit 115 using a predetermined forgetting factor μ. The transfer function matrix Hnew is calculated.
Next, in the transfer function matrix management circuit 115, the transfer function matrix update control circuit 9 overwrites the obtained transfer function matrix Hnew on the transfer function matrix Hnow to obtain a new transfer function matrix Hnow.

The forgetting factor μ used in the calculation of the new transfer function matrix generation circuit 8 is desirably set in proportion to channel estimation, that is, the estimation accuracy of the transfer function matrix.
Also, the estimation accuracy of the transfer function matrix is expected to vary depending on the determinant of the virtual transmission signal matrix (transmission signal replica matrix), but there are many to calculate the determinant of the 4 × 4 matrix directly. A multiplication circuit is required, and the calculation load increases with the circuit scale.
However, in the case of the above-described embodiment, the determinant of the virtual transmission signal matrix is the absolute value of the virtual transmission signal sequence vector Tx {t1, t2, t3, t4} whose value is four signal points (the magnitude of the vector). ) Product.

That is, by using the forgetting factor μ (| t1 * t2 * t3 * t4 |) as a function of the absolute value | t1 * t2 * t3 * t4 |, the expression (6) is changed to the expression (10) shown below. By replacing it, the amount of calculation at that time can be suppressed.
Hnew = (1-μ (| t1 * t2 * t3 * t4 |)) · Hnow
+ Μ (| t1 * t2 * t3 * t4 |) · Hest (10)

In the embodiment described above, MIMO reception signal detection (transmission signal estimation processing) is shown as the signal detection circuit 116. However, the transmission signal (MLD, ZF, etc.) shown in the conventional example is estimated by a known method. Using the output of the error-corrected data), the transfer function matrix (channel) tracking process is performed by performing the above-described process.
In addition, the transfer function matrix update control circuit 9 may replace the transfer function Hest with a simply generated transfer function Hnew, or update it by adding some condition (for example, the strength of the received signal level). Also good.

Furthermore, in the above-described embodiment, the case where the multiplexing number in the MIMO transmission is 4 is described as an example, and the test matrix U is described using the matrix shown on the right side of Equation (7). The characteristics of this matrix are the four column vectors (1,1,1,1) ^t , (1, -1, -1,1) ^t , (1,1, -1, -1) ^t constituting the matrix , (1, -1,1, -1) ^t are orthogonal (the inner product of any two vectors is zero). There are many matrices having such characteristics. For example, even when each column is replaced as appropriate, or when each column is multiplied by an appropriate coefficient, the matrix has the same characteristics.
In particular, when considering a matrix in which all elements are multiplied by 1/2 as a coefficient, this matrix has the original matrix and its Hermitian conjugate matrix (transposed complex conjugate) in addition to the above properties. It also has the property that the product of is a unit matrix. A matrix having such characteristics is called a unitary matrix. There are many test matrix candidates having the above properties, and any matrix among them can be used as the test matrix.
The above-described embodiments are all illustrative of the present invention and are not limited to the present invention, and the present invention can be implemented in various other variations and modifications.
Therefore, the scope of the present invention is defined only by the claims and their equivalents.

  Note that a program for realizing the function of the receiving unit in FIG. 1 is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed, whereby data of the received data is received. Regeneration may be performed. Here, the “computer system” includes an OS and hardware such as peripheral devices. The “computer system” includes a WWW system having a homepage providing environment (or display environment). The “computer-readable recording medium” refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, and a CD-ROM, and a storage device such as a hard disk built in the computer system. Further, the “computer-readable recording medium” refers to a volatile memory (RAM) in a computer system that becomes a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. In addition, those holding programs for a certain period of time are also included.

  The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, and what is called a difference file (difference program) may be sufficient.

It is a block diagram which shows the example of 1 structure of the receiving part of the receiving station by one Embodiment of this invention. It is a block diagram which shows the structure of the transmission part of the transmission station to which a MIMO technique is applied. It is a block diagram which shows the structure of the receiving part of the receiving station in a prior art.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Diagonal matrix production | generation circuit 2 ... Diagonal inverse matrix production | generation circuit 3 ... 1st matrix multiplication circuit 4 ... Test matrix memory circuit 5 ... Received signal conversion circuit 6 ... Replica Matrix generation circuit 7... Second matrix operation circuit 8... New transfer function matrix generation circuit 9... Transfer function matrix update control circuit 101... Data division circuit 102-1, 102-2, 102 − N: Preamble providing circuits 103-1, 103-2, 103-N: Modulation circuits 104-1, 104-2, 104-N: Radio units 105-1, 105-2, 105-M ... Transmission antennas 111-1, 111-2, 111-M ... Reception antennas 112-1, 112-2, 112-M ... Radio unit 113 ... Channel estimation circuit 114 ... Reception signal Management circuit 115 ... transfer function matrix management circuit 16 ... signal detecting circuit 117 ... data combining circuit

Claims (7)

A transmitting station having N (N ≧ 2, N is an integer) transmitting antennas that multiplex and transmit a plurality of signal sequences in space on the same frequency channel; In a wireless communication system capable of MIMO (Multiple Input Multiple Output) communication configured by a receiving station provided with M receiving antennas (M ≧ 1, M is an integer) that performs reception processing by separating into signal sequences,
The transmitting station is
User data dividing means for dividing user data into N systems;
A signal sequence generation means for generating a signal sequence of N systems by giving individual known pattern signals to the data divided into the N systems;
Signal transmitting means for simultaneously superimposing and transmitting the signal sequence at the same frequency using N transmitting antennas,
The receiving station is
Signal receiving means for individually receiving radio signals using the M receiving antennas;
M × N sets of transfer functions h _j, between the i-th antenna of the transmitting antennas and the j-th antenna of the receiving antennas, using a signal of a known pattern given to the received signal as a reference signal _. transfer function acquisition means for acquiring _i ;
A matrix of M rows and N columns, that is, a transfer function matrix having the transfer function h _{j, i} as the (j, i) component, that is, a transmission function matrix N, a transmission signal sequence vector T of N rows and 1 column, When a received signal in units of symbols received by M antennas is expressed as a received signal column vector R of M rows and 1 column,
Transmission signal estimation means for obtaining an estimated transmission signal sequence vector T ′ which is an estimated value of the transmission signal sequence vector T transmitted by the transmission station using the transfer function matrix H and the received signal sequence vector R for each symbol. When,
A test matrix which is a matrix of N rows and N columns, and all the values of the components of the predetermined k-th column (1 ≦ k ≦ N: k is an integer) are constants C (C is an arbitrary real number or complex number). Test matrix storage means for storing U and an inverse matrix U ⁻¹ of the matrix;
Diagonal matrix generating means for generating an N-by-N square diagonal matrix D in which N components of the estimated transmission signal sequence vector T ′ are diagonal components and all non-diagonal components are zero;
Inverse matrix obtaining means for obtaining an inverse matrix D ⁻¹ of the square diagonal matrix D;
First matrix multiplication means for obtaining a product of the transfer function matrix H, the square diagonal matrix D and the test matrix U, that is, a matrix H · D · U of M rows and N columns;
A replica in which each component of the received signal sequence vector R is converted to a constant C times to generate a converted received signal sequence vector R ′, and the kth row of the matrix of M rows and N columns is replaced with the converted received signal sequence vector R ′. Replica matrix generation means for generating a matrix H _rep ;
Second matrix multiplication for obtaining a product of the replica matrix H _rep , the inverse matrix U ^{−1 of the} test matrix, and the square diagonal matrix D ⁻¹ , that is, a matrix H _rep · U ⁻¹ · D ⁻¹ of M rows and N columns. Means,
New transfer function matrix generating means for generating a new transfer function matrix as a linear combination of the matrix and the transfer function matrix;
A wireless communication system, comprising: transfer function updating means for updating a transfer function matrix that has been used for reception processing up to that time to the new transfer function matrix. The wireless communication device according to claim 1,
In the test matrix U, each column vector constituting the matrix is orthogonal to each other. The wireless communication apparatus according to claim 2, wherein
The test matrix U is a unitary matrix or a matrix obtained by multiplying each component of the unitary matrix by a constant. In the wireless communication apparatus according to any one of claims 1 to 3,
The new transfer function matrix generation means uses a forgetting factor μ (0 <μ ≦ 1: μ is a real number), and (1−μ) times the transfer function H and the matrix H _rep · U ⁻¹ · D ⁻¹ . A wireless communication system characterized by generating a new transfer function matrix by linearly combining μ times. The wireless communication device according to claim 4, wherein
The forgetting factor μ is dynamically changed in proportion to the absolute value of the estimated transmission signal sequence vector T ′. Transmitted by a MIMO (Multiple Input Multiple Output) communication system from a transmission station equipped with N (N ≧ 2, N is an integer) transmission antennas that multiplex and transmit a plurality of signal sequences on the same frequency channel. A receiving apparatus including M (M ≧ 1, M is an integer) receiving antennas that receive received radio signals and perform reception processing by separating them into a plurality of signal sequences;
Signal receiving means for individually receiving radio signals using the M receiving antennas;
M × N sets of transfer functions h _j, between the i-th antenna of the transmitting antennas and the j-th antenna of the receiving antennas, using a signal of a known pattern given to the received signal as a reference signal _. transfer function acquisition means for acquiring _i ;
A matrix of M rows and N columns, that is, a transfer function matrix having the transfer function h _{j, i} as the (j, i) component, that is, a transmission function matrix N, a transmission signal sequence vector T of N rows and 1 column, When a received signal in units of symbols received by M antennas is expressed as a received signal column vector R of M rows and 1 column,
Transmission signal estimation means for obtaining an estimated transmission signal sequence vector T ′ which is an estimated value of the transmission signal sequence vector T transmitted by the transmission station using the transfer function matrix H and the received signal sequence vector R for each symbol. When,
A test matrix which is a matrix of N rows and N columns, and all the values of the components of the predetermined k-th column (1 ≦ k ≦ N: k is an integer) are constants C (C is an arbitrary real number or complex number). Test matrix storage means for storing U and an inverse matrix U ⁻¹ of the matrix;
Diagonal matrix generating means for generating an N-by-N square diagonal matrix D in which N components of the estimated transmission signal sequence vector T ′ are diagonal components and all non-diagonal components are zero;
Inverse matrix obtaining means for obtaining an inverse matrix D ⁻¹ of the square diagonal matrix D;
First matrix multiplication means for obtaining a product of the transfer function matrix H, the square diagonal matrix D and the test matrix U, that is, a matrix H · D · U of M rows and N columns;
A replica in which each component of the received signal sequence vector R is converted to a constant C times to generate a converted received signal sequence vector R ′, and the kth row of the matrix of M rows and N columns is replaced with the converted received signal sequence vector R ′. Replica matrix generation means for generating a matrix H _rep ;
Second matrix multiplication for obtaining a product of the replica matrix H _rep , the inverse matrix U ^{−1 of the} test matrix, and the square diagonal matrix D ⁻¹ , that is, a matrix H _rep · U ⁻¹ · D ⁻¹ of M rows and N columns. Means,
New transfer function matrix generating means for generating a new transfer function matrix as a linear combination of the matrix and the transfer function matrix;
A transfer device comprising: transfer function updating means for updating a transfer function matrix used in the previous reception process to the new transfer function matrix. Transmitted by a MIMO (Multiple Input Multiple Output) communication system from a transmission station equipped with N (N ≧ 2, N is an integer) transmission antennas that multiplex and transmit a plurality of signal sequences on the same frequency channel. A reception method of receiving the received radio signal by M reception antennas (M ≧ 1, M is an integer) that performs reception processing by separating the signal into a plurality of signal sequences,
A signal receiving process of individually receiving radio signals using the M receiving antennas;
M × N sets of transfer functions h _j, between the i-th antenna of the transmitting antennas and the j-th antenna of the receiving antennas, using a signal of a known pattern given to the received signal as a reference signal _. a transfer function acquisition process for acquiring _i ,
A matrix of M rows and N columns, that is, a transfer function matrix having the transfer function h _{j, i} as the (j, i) component, that is, a transmission function matrix N, a transmission signal sequence vector T of N rows and 1 column, When a received signal in units of symbols received by M antennas is expressed as a received signal column vector R of M rows and 1 column,
A transmission signal estimation process for obtaining an estimated transmission signal sequence vector T ′ that is an estimated value of the transmission signal sequence vector T transmitted by the transmission station using the transfer function matrix H and the received signal sequence vector R for each symbol. When,
A test matrix which is a matrix of N rows and N columns, and all the values of the components of the predetermined k-th column (1 ≦ k ≦ N: k is an integer) are constants C (C is an arbitrary real number or complex number). A test matrix storage process for storing U and an inverse matrix U ⁻¹ of the matrix in a test matrix storage means;
A diagonal matrix generating step of generating an N-by-N square diagonal matrix D in which N components of the estimated transmission signal sequence vector T ′ are diagonal components and all non-diagonal components are zero;
Inverse matrix obtaining means for obtaining an inverse matrix D ⁻¹ of the square diagonal matrix D;
A first matrix multiplication process for obtaining a product of the transfer function matrix H, the square diagonal matrix D, and the test matrix U, that is, a matrix H · D · U of M rows and N columns;
A replica in which each component of the received signal sequence vector R is converted to a constant C times to generate a converted received signal sequence vector R ′, and the kth row of the matrix of M rows and N columns is replaced with the converted received signal sequence vector R ′. A replica matrix generation process for generating a matrix H _rep ;
Second matrix multiplication for obtaining a product of the replica matrix H _rep , the inverse matrix U ^{−1 of the} test matrix, and the square diagonal matrix D ⁻¹ , that is, a matrix H _rep · U ⁻¹ · D ⁻¹ of M rows and N columns. Process,
A new transfer function matrix generation process for generating a new transfer function matrix as a linear combination of the matrix and the transfer function matrix;
And a transfer function updating step of updating the transfer function matrix used in the previous reception process to the new transfer function matrix.

Images