JP5456638B2 - Wireless communication method and wireless communication system - Google Patents

Description

  The present invention relates to a radio communication method and a radio communication system in which each base station performs signal transmission cooperatively in an outdoor radio access system in which a plurality of base stations perform signal transmission using the same frequency channel.

  In order to provide a broadband Internet connection service, an optical line is spreading. However, laying an optical line is very expensive, and it is difficult to lay it unless a certain number of users can be expected. Therefore, in order to provide a broadband Internet connection service with reduced facility costs, a method of using a wireless line at a location closest to the user (last one hop) has been studied.

  As this last one-hop wireless line, it is ideal to use a microwave band in which communication is possible without a line of sight between the base station on the network side and the terminal station on the user's home side. However, in the present situation where frequency resources are being exhausted, it is difficult for a single operator to monopolize many frequency channels in the microwave band. On the other hand, since it is difficult to cover a wide service area with one base station, when expanding the service area in one area, one base station covers a service area called a cell in a circular shape, and It has been dealt with by spreading cells continuously. At this time, if adjacent cells use the same frequency channel, inter-cell interference generally occurs, and the characteristics are greatly deteriorated. Usually, in order to solve such a problem, a plurality of frequency channels are used, and the frequency interval is assigned repeatedly, thereby isolating the cell interval of the same frequency channel.

  As described above, it is possible to suppress mutual interference between cells of the same frequency channel by repeatedly assigning a plurality of frequency channels. However, if there are few available frequency channels, the interference is not necessarily reduced to a sufficient level. There is no way to suppress it. As a method for solving such a problem, a method using an intercell interference canceller for suppressing mutual interference between cells as described in Non-Patent Document 1 has been proposed.

FIG. 3 shows a configuration example of a wireless communication system using an inter-cell interference canceller.
In FIG. 3, 101 is a control station, 102-1 to 102-5 are remote base stations, 103-1 to 103-5 are terminal stations, 104-1 to 104-5 are cells using the same frequency channel, and 105 is wired. Represents a transmission line. The cells 104-1 to 104-5 using the same frequency are separated from each other by a certain distance due to repeated use of a plurality of frequency channels. Although the cells other than the cell of the frequency channel of interest are not described here, there are actually cells using other frequency channels. The remote base stations 102-1 to 102-5 installed in the cells 104-1 to 104-5 are connected to the control station 101 via the wired transmission path 105. The control station 101 collectively manages wireless communication between each of the remote base stations 102-1 to 102-5 and the terminal stations 103-1 to 103-5 subordinate thereto, and performs various signal processing.

  The remote base stations 102-1 to 102-5 and the terminal stations 103-1 to 103-5 are provided with a plurality of antennas in the figure, and are configured so as to configure a MIMO channel for each of the cells 104-1 to 104-5. Although shown, each may constitute a SISO channel with one antenna. Furthermore, a configuration may be adopted in which a plurality of remote base stations are provided in each cell, and a plurality of antennas are provided as a whole. Further, the terminal stations 103-1 to 103-5 may exist in a plurality of stations in the cell, and may be configured to realize multiuser MIMO communication in which communication is performed in parallel using the same frequency channel at the same time.

  It is assumed that the control station 101 can acquire channel information of the MIMO channel between the remote base stations 102-1 to 102-5 and the terminal stations 103-1 to 103-5 by some method. Since this channel information acquisition method is a general technique discussed in various documents, the details are omitted here because the channel information is known.

Next, the overall channel matrix H _all in the wireless communication system is defined as follows.

Here, N represents the total number of cells using the same frequency channel constituting the wireless communication system. Further, each component H _{i, j} constituting the channel matrix H _all itself constitutes a matrix. For example, in the example of FIG. 5, N = 5 because it is composed of five cells, and the diagonal components H _1,1 , H _{2,2 ,.} -1 to 104-5 represents a MIMO channel between the base stations 102-1 to 102-5 and the terminal stations 103-1 to 103-5. H _i, j for i ≠ j, which is a non-diagonal component, represents a channel matrix corresponding to interference from the base station 102-j in the j-th cell to the terminal station 103-i in the i-th cell. Since this matrix H _{i, j} is different for each terminal station that is a communication partner, it should be marked with a subscript corresponding to that terminal station, but only the terminal that is to be communicated at a certain moment. Here, for convenience of explanation, subscripts corresponding to terminal stations are omitted.

Here, when the norm of the off-diagonal matrix is sufficiently smaller than the norm of the diagonal matrix, that is, when the following condition is satisfied, the interference canceller functions effectively.

In the following description, the matrix H _{d in} which only the diagonal terms of the channel matrix H _all are extracted and zeros are inserted, and the matrix H _{nd in} which only the non-diagonal terms are extracted and the diagonal terms are zero-inserted are as follows: Define as follows.

Further, when the transmission weight that the remote base station multiplies the transmission signal as necessary in the i-th cell is W _{i, i} , the entire transmission weight matrix in which this partial matrix is arranged in a diagonal term is as follows: Define.

Similarly, if the transmission information from the remote base station in the i-th cell is S _i , the signal received at the i-th cell terminal station is R _i , and the noise signal at the i-th cell terminal station is n _i , As a whole, it can be expressed as follows.

The signal T in Expression (10) is a signal obtained by multiplying transmission information by a transmission weight, and is positioned as a precoded signal that is actually transmitted from the transmission side. Here, since the transmission weight matrix W _{i, i} is a transmission weight calculated by ignoring interference from cells other than its own cell, the inter-cell interference signal is mixed in Equation (6). ing. In the interference canceller described in Non-Patent Document 1, it is estimated how the sum of interference signals from other cells is received at the terminal station, and a replica of the signal with the opposite sign of the estimated signal is used as the remote base station. Is generated and added to the original signal and transmitted. A specific transmission signal T ′ in consideration of the interference replica signal is given by the following equation.

Here, equation (11) can be understood as calculating the following converted transmission weight W ′ in place of the previous transmission weight W and multiplying the transmission information S by W ′ to obtain the transmission signal.

  FIG. 4 shows a downlink transmission signal calculation processing procedure in a conventional wireless communication method. FIG. 4 (a) is a channel information acquisition process periodically performed between the remote base station and each terminal station, FIG. 4 (b) is a previous transmission signal calculation preparation process for performing actual transmission process, and FIG. 4 (c). Represents a transmission signal calculation process in symbol units when transmitting each bit string.

In FIG. 4 (a), the propagation channel between the remote base station and each terminal station generally changes with time. Therefore, each channel information is periodically acquired at a predetermined cycle. Specifically, when processing is started (S101), channel information is acquired (S102), and it is recorded as a partial channel matrix H _{i, j} of Equation (3) and Equation (4) (S103). The process ends (S104). Since the partial channel matrix H _{i, j} is originally different for each terminal station, a subscript corresponding to the terminal station should be given, but is omitted here for convenience of explanation.

In FIG. 4 (b), when the transmission opportunity is obtained and the process is started (S111), the terminal station of the communication partner is selected (S112), and the partial channel matrix H _{i, j} related to the terminal station stored in the memory is selected. Is read (S113). Further calculates versus partial channel matrix H _i corner _components, transmission weight W _i for _i, a _i (S114). The transmission weight here is, for example, a transmission weight for eigenmode transmission in MIMO transmission or a transmission weight for null formation for inter-terminal-station interference suppression in multi-user MIMO. It doesn't matter. In addition, a simple unit matrix may be used without particularly defining a transmission weight. In this case, the transmission weight calculation process (S114) has no meaning effectively. Next, a converted transmission weight W ′ represented by the equation (12) is calculated for the transmission weight W having the transmission weights Wi _{, i} as diagonal components (S115), and the process is terminated (S116).

  In FIG. 4 (c), when the processing is started as the calculation processing of the transmission signal that is actually precoded based on the bit string (S121), the transmission information S is input (S122), and the transmission information S according to the equation (11). Is multiplied by the converted transmission weight W ′ to calculate the transmission signal T ′ (S123), and the processing is terminated to determine the transmission signal T ′ (S124). The transmission information S is processed as a signal vector obtained by synthesizing the components for each cell according to the equation (7).

  Here, in order for this interference canceller to function effectively, the condition of Equation (2) must be satisfied, and if the repetition was originally performed with a small number of frequency channels, inter-cell interference should be sufficiently suppressed. I could not.

Therefore, in an environment where the condition of equation (2) is insufficient, it is effective to further increase the estimation accuracy of the interference replica signal. The replica signal for canceling the interference signal shown in the previous equation (11) ignores the interference signal from an adjacent cell that can be an interference source for the target cell for which this replica signal is obtained, and only the single cell is used. It was obtained on the premise of the transmission signal in consideration. If this is expanded, the replica signal for canceling the signal is calculated again, and the signal is subtracted on the assumption that the signal given by Equation (11) is transmitted from a cell that can be an adjacent interference source. This makes it possible to increase the accuracy of approximation. A specific transmission signal T ″ in consideration of the interference replica signal in this case is given by the following equation.

In general, N ³ multiplications are required to multiply N × N matrices. As described in Non-Patent Document 1, as the service area as a whole becomes wide, and the number of remote base station antennas in the entire area increases, the number of multiplications increases in proportion to N ³ . That is, although it is possible to increase the accuracy of approximation, application in an environment in which the size of the entire matrix _Hall is increased is impractical because the circuit scale increases and the calculation load increases. It was.

  In contrast to the above-described downlink signal processing, uplink signal processing will be briefly described. In the uplink, each terminal station does not have sufficient knowledge of channel information with remote base stations in neighboring cells, and each terminal station in multiple cells communicates in a coordinated manner. Therefore, it is impossible for each remote base station to perform coordinated transmission that suppresses the interference / interference between the cells in each remote base station.

Therefore, if the transmission information of each cell in the uplink is defined in the same way as Equation (7), since all terminal stations transmit signals without multiplying a special transmission weight, Equation ( 6) can be rewritten as follows.

Here, H ′ _all is the entire channel matrix corresponding to the uplink. Considering that the size of this matrix is very large, it is impossible to apply a maximum likelihood detection (MLD) method that exhibits good characteristics in normal MIMO communication. Therefore, signal processing using a ZF (Zero Forcing) method, a minimum square error (MMSE) method, or the like is assumed. Processing contents in the case of the simplest ZF method will be described below.

First, if the matrix H ′ _all is a matrix of size N × M, signal separation cannot be performed unless the total number N of receiving antennas is larger than the number M of transmission signal systems, so that the relationship of N ≧ M is inevitably established. . If N = M, signal detection processing can be performed by multiplying Expression (14) by an inverse matrix of matrix H ′ _all from the left side. However, in general, N = M is not always true, and thus the matrix H ′ _all itself does not always have an inverse matrix. In this case, the matrix H 'is a pseudo-inverse matrix of _all the matrix ^{_{(H' all H H 'all}} ) -1 H' all H
Is used. Here, the matrix H ′ _all ^H H ′ _all is a square matrix of size M × M, and if the rank of the matrix H ′ _all is M, the matrix H ′ _all ^H H ′ _all generally has an inverse matrix. Exists. Therefore, by multiplying Equation (14) by the pseudo inverse matrix from the left side, the following equation is obtained.

  The second term on the right-hand side is a signal obtained by synthesizing each component of the noise vector n with a weight of a pseudo inverse matrix, and is generally sufficiently smaller than the size of the transmission information S. Based on the signal obtained by performing the soft decision (with error correction processing) processing, transmission information transmitted on the terminal station side is estimated (detection of the received signal).

  FIG. 5 shows an uplink received signal estimation processing procedure in a conventional wireless communication method. Fig. 5 (a) is a channel information acquisition process periodically performed between the remote base station and each terminal station, Fig. 5 (b) is a preparatory process for the reception signal estimation process, and Fig. 5 (c) is a data process. The received signal estimation processing in symbol units when receiving a bit string is shown.

In FIG. 5 (a), the propagation channel between the remote base station and each terminal station generally changes with time. Therefore, each channel information is periodically acquired at a predetermined cycle. Specifically, when the process is started (S131), channel information is acquired (S132), recorded as a partial channel matrix H ′ _{i, j} (S133), and the process ends (S134). Originally, since the partial channel matrix H ′ _{i, j} is different for each terminal station, a subscript corresponding to the terminal station should be given, but is omitted here for convenience of explanation.

Next, in FIG. 5 (b), when processing is started by data reception (S141), channel information is acquired in each cell (S142), and the submatrix H ′ between cells acquired in FIG. 5 (a) is obtained. _i, reads _j (S143), they synthesized to create a whole matrix H _'all in (S144). For this entire channel matrix, a pseudo inverse matrix is calculated by equation (15) (S145), the process is terminated (S146), and the received signal estimation process for information (bit string) following the received data is continued.

In general, since a preamble signal for channel estimation is added to the head region of the data to be received, it is transmitted from the terminal in the focused cell corresponding to the diagonal term of the partial channel matrix in step S142. The partial channel matrix H ′ _{i, i} when received by the base station in this cell can be acquired. However, the partial channel matrix H ′ _{i, j} corresponding to the off-diagonal term depends on the system and cannot always be acquired. In the explanation of FIG. 5 (b), the partial channel matrix H ′ _{i, j} corresponding to the off-diagonal term is acquired at a different opportunity from the actual data reception as explained in FIG. 5 (a). The case where the information recorded in step S133 in FIG. 5A is read and used in step S143 has been described as an example. However, if the partial channel matrix H ′ _{i, j} corresponding to the off-diagonal term can be acquired at the time of data reception, the series of processing shown in FIG. 5 (a) can be omitted. All diagonal terms and non-diagonal terms are acquired, and processing S143 is omitted.

Next, in FIG. 5 (c), when starting the reception signal estimation processing the bit sequence of the data on a symbol basis (S151), a reception signal of the i-th cell of the symbol units R _i obtained in all the cells (S152), According to the equation (15), the pseudo inverse matrix is multiplied from the left side of the entire received signal vector R to calculate the estimated signal S (S153), and the received signal estimating process is terminated (S154).

  In order to suppress estimation errors due to noise components, etc., the processing S153 includes a signal hard decision processing, a soft decision processing including error correction, and the like. Omitted.

Kazuteru Maruta et al., “Improvement efficiency of frequency utilization using inter-cluster interference canceller in multi-user MIMO distributed antenna system”, IEICE Technical Report RCS2009-231, pp.139-144, January 2010

The interference canceller described in Non-Patent Document 1 is a technique related only to signal processing related to the downlink, and there was no mention of signal processing related to the uplink. Although it is possible in principle to perform signal processing by receiving signals transmitted by a large number of terminal stations, here, as shown in Equation (15), matrix multiplication or inverse matrix is performed on the matrix. With computation. Since this matrix is composed of the entire matrix for all service areas, it has a very large size. For example, if the size is 100 × 100, each multiplication and inverse matrix operation requires 10 ⁶ multiplications. In reality, it is also common for the matrix size to be larger than this, and in order to process a huge amount of computation, it is assumed that hardware with an unrealistic circuit scale is assumed, or software with a huge processing delay Processing is inevitable.

  The present invention establishes an uplink interference canceller technology that can be realized with a realistic amount of computation and a realistic circuit scale, and can effectively use frequency resources by suppressing the number of frequency channels for frequency repetition. An object is to provide a wireless communication method and a wireless communication system.

In the present invention, one or more base stations are arranged in each of a plurality of cells, each base station is connected to a network via a wired line, and each base station communicates with a terminal station in the cell via a wireless line. A wireless communication method, the first step of acquiring channel information between a base station and a terminal station in the same cell and between different cells, and the channel information acquired in the first step in the i-th cell A second step of managing as a channel matrix H ′ _{i, j} between the base station and the terminal station in the j-th cell, and a reception weight matrix W ′ _{i, i} between the base station and the terminal station in the same cell. Are obtained based on the channel matrix, and for the channel matrix,
G _{i, j} = W ′ _{i, i} H ′ _{i, j} (W ′ _{j, j} H ′ _{j, j} ) ⁻¹
The third step of obtaining the matrix G _{i, j} and the reception information in the i-th cell is R _i, and the 0th-order estimated signal in the i-th cell is multiplied by the reception weight matrix W ′ _{i, i.} A fourth step of obtaining S _i ^[0], and the a-th order in the j-th cell where j ≠ i calculated in all or part of the surrounding cells that communicate using the same frequency channel as the i-th cell. a fifth step of acquiring the estimated signal S _j ^[a], the estimated signal S _j ^[a] a matrix G _i, the result of integration _j take the sum for a given j, and the result estimated signal A sixth step of calculating S _i ^{[a + 1]} from the a + 1-th order estimated signal in the i-th cell by subtracting from S _i ^[0] is repeatedly performed to generate an estimated signal. .

It differs from the conventional system in that the third step, the fourth step, the fifth step, and the sixth step that is repeatedly performed are performed. Accordingly, even when the matrix size of the entire matrix H _'all is very large, the matrix G _i is a common small matrix for the value of each _a, while using the _j, the a next estimation signal S _i ^[ By calculating ^a] with a recurrence formula, it is possible to obtain a solution of a higher-order estimated signal while suppressing the amount of calculation.

  In addition, one or more base stations are arranged in each of a plurality of cells, each base station is connected to a network via a wired line, and each base station communicates with a terminal station in the cell via the radio line. It is the structure which communicates by.

  As a result, a method for reducing the circuit scale of each facility and creating a realistic circuit configuration by distributing the processing of the entire service area to a plurality of facilities without consolidating the processing to a single base station or control station. Can be given.

  According to the present invention, in an uplink in which multiple remote base stations receive signals transmitted by multiple terminal stations in an environment in which mutual interference / interference from neighboring cells using the same frequency channel exists, Communicating with improved signal-to-interference ratio (SIR) by suppressing mutual interference / interference from surrounding cells by signal processing of received signals from antennas at remote base stations be able to. As a result, it is possible to reduce the number of frequency channels required for frequency repetition and to effectively use frequency resources.

  Furthermore, in the prior art, the amount of calculation required in the present invention can be greatly suppressed with respect to the signal processing of an interference canceller that involves a huge amount of calculation on an unrealistic scale.

  Also, it is not necessary to consolidate all signal processing itself in one control station, and processing can be executed in a distributed control station or base station. As a result, the calculation amount of individual control stations or base stations is reduced to the service area. It becomes possible to limit the amount of computation to a practical amount that does not depend on the total number of cells, and it is possible to extend to an extremely wide service area in the limit.

It is a flowchart which shows the received signal estimation processing procedure in the radio | wireless communication method of this invention. It is a figure which shows the structural example of the radio | wireless communications system using the inter-cell interference canceller in the Example of this invention. It is a figure which shows the structural example of the radio | wireless communications system using the intercell interference canceller. It is a flowchart which shows the transmission signal calculation process procedure of the downlink in the conventional radio | wireless communication method. The It is a flowchart which shows the received signal estimation processing procedure of the uplink in the conventional radio | wireless communication method.

First, the principle of the present invention will be described.
Here, the channel matrix for the uplink is redefined.

Here, W ′ _{i, i} in each term of W ′ in Expression (19) is a reception weight for H ′ _{i, i} in Expression (17). For example, in the ZF method, W ′ _{i, i} = H ' _{i, i} ^-1
Or W ′ _{i, i} = (H ′ _{i, i} ^H H ′ _{i, i} ) ⁻¹ H ′ _{i, i} ^H
Given in. Further, it may be calculated by any other method such as a weight based on the least mean square error (MMSE) standard. Multiply this matrix from the left side of both sides of Equation (14).

Further, the calculation proceeds as follows.

Here, if the noise term of the third term on the right side of the equation (20) is ignored, the following relational expression can be obtained.

If this is described separately for each line, the i-th line becomes the following relational expression.

Here, W ′ _{i, i} H ′ _{i, i} multiplied by S _i on the left side is a diagonal matrix. This is because information on the signal when there is no inter-cell interference is obtained by estimating and subtracting the signal that is an interference component from W ′ _{i, i} R _{i obtained} by multiplying the received signal by the reception weight (the left side of Expression (24) ) Is required. If transmission information S _i satisfying Expression (24) can be obtained for all i and j, it can be regarded as an approximate solution of transmission information S. Therefore, the following processing is performed.

Equation (27) is ^a recurrence formula for the transmission information S _i ^[a] . If the initial value is given by Equation (25) and then repeated calculation is performed, the conditional equation of Equation (2) is satisfied. If so, it will have a convergent solution. Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 shows a received signal estimation processing procedure in the wireless communication method of the present invention. Fig. 1 (a) is a channel information acquisition process periodically performed between the remote base station and each terminal station, Fig. 1 (b) is a preparatory process for the reception signal estimation process, and Fig. 1 (c) is a data process. The received signal estimation process in symbol units when each bit string is received is shown.

In FIG. 1 (a), the propagation channel between the remote base station and each terminal station generally changes with time. Therefore, each channel information is periodically acquired at a predetermined cycle. Specifically, when the process is started (S1), channel information is acquired (S2), and is recorded as a partial channel matrix H ′ _{i, j} of Expression (16) (S3), and the process is terminated (S4). ). Originally, since the partial channel matrix H ′ _{i, j} is different for each terminal station, a subscript corresponding to the terminal station should be given, but is omitted here for convenience of explanation.

In FIG. 1 (b), when data is received and processing is started (S11), channel information with a terminal station as a communication partner in the focused cell is acquired (S12), and a partial channel matrix H ′ _{i, i} (S13). Further, the reception weight matrix W ′ _{i, i} between the base station and the terminal station in the same cell is acquired based on the partial channel matrix H ′ _{i, i,} and the matrix G _{i, j} is calculated according to equation (26). (S14), the process ends (S15). Then, the received signal estimation process for information (bit string) following the received data is continued.

Here, the same description is given in the prior art in FIG. 5, but since a preamble signal for channel estimation is added to the head region of received data, it corresponds to the diagonal term of the partial channel matrix in step S12. However, the partial channel matrix H ′ _{i, i} when transmitted from the terminal in the focused cell and received by the base station in the cell can be acquired. However, the partial channel matrix H ′ _{i, j} corresponding to the off-diagonal term depends on the system and cannot always be acquired. In the explanation in FIG. 1 (b), the partial channel matrix H ′ _{i, j} corresponding to the off-diagonal term is acquired at a different opportunity from the actual data reception as explained in FIG. 1 (a). The case where the information recorded in step S3 in FIG. 1A is read out and used in step S13 has been described as an example. However, if the partial channel matrix H ′ _{i, j} corresponding to the off-diagonal term can be acquired at the time of data reception, the series of processing shown in FIG. 1 (a) can be omitted, and the processing in step S12 All diagonal terms and non-diagonal terms are acquired, and the process S13 is omitted.

In FIG. 1 (c), when a bit string of data is received in symbol units and the received signal estimation process is started (S21), the received signal of the i-th cell in symbol units is set to R _i (S22), according to equation (25). S _j ^[0] is calculated by the 0th-order signal detection processing of the focused cell (S23). Next, the counter value a is reset to zero (S24), the signal S _i ^[a] of the neighboring cell is acquired (S25), and the estimated signal S _i ^{[a + 1] of} the cell focused on a + 1 according to the equation (27) ^. Is calculated (S26). Further, 1 is added to the counter value a (S27), and it is determined whether or not a predetermined threshold value b is exceeded (S28). If Yes in step S28, the process ends and the b-th order estimated signal S _i ^[b] is determined (S29). On the other hand, if No in step S28, the process returns to step S25, and the processes from step S25 to step S28 are repeated.

  In step S29, the signal estimation process is completed. Naturally, in order to suppress an estimation error due to a noise component or the like, a signal hard decision process, a soft decision process including error correction, and the like may be included. Since these processes are general techniques, description thereof is omitted here.

Here, supplementing step S25, the peripheral cells here are cells to be summed by Σ in equation (27), and specifically, a predetermined level at which mutual addition / interference cannot be ignored. This is the above cell. Since this condition is usually determined by station design, it is generally fixed. However, the degree of successive interference is investigated, and cells that cannot be ignored at that time are dynamically managed. It doesn't matter. Further, since only the estimated signal S _i ^[a] obtained as an a-th order approximate solution needs to be exchanged between neighboring cells, the amount of information notified to each other is limited.

  In addition, when the process from step S25 to step S28 is repeatedly performed by configuring a circuit on hardware, the same circuit is repeatedly used, so the amount of calculation slightly increases in terms of the number of calculations. The circuit scale does not depend on the number of repetitions of the processing from step S25 to step S28, and can be kept constant.

  FIG. 2 shows a configuration example of a radio communication system using an inter-cell interference canceller in an embodiment of the present invention.

  In FIG. 2, 1-1 to 1-2 are control stations, 2-1 to 2-5 are remote base stations, 3-1 to 3-5 are terminal stations, and 4-1 to 4-5 are the same frequency channel. Cells used, 5-1 to 5-3, represent wired transmission lines.

  The difference between the configuration of this embodiment and the conventional configuration shown in FIG. 3 is that remote base stations 2-1 to 2-5 in all cells 4-1 to 4-5 are connected to a single control station. Rather, it is connected to any one of the plurality of dispersed control stations 1-1 to 1-2 via the wired transmission paths 5-1 to 5-2, and the plurality of dispersed control stations 1-1 to 1- 2 is also connected through a wired transmission line 5-3.

For example, in the cell 4-1 of interest, the cells 4-2 and 4-5 that cannot be ignored or interfered with each other are the cells 4-2 and 4-5. In this case, in the control station 1-1 to which the remote base station 2-1 in the cell 4-1 is connected, between the cell 4-1 and between the cell 4-1 and the cell 4-2 / cell 4-5 Channel information is acquired, and the matrix G _{i, j} shown in Equation (26) is calculated. Further, when actually receiving the signal, S _i ^[a] in the equation (28) is replaced with the control station 1-1 to which the remote base station 2-2 of the cell 4-2 is connected (actually, the same control station). Therefore, the information is not required to be transmitted) and is acquired from the control station 1-2 to which the remote base station 2-5 of the cell 4-5 is connected (step S25 in FIG. 1 (c)). This information exchange is repeated while repeating the processing of steps S25 to S28 in FIG. 1 (c), and finally the estimated signal S _i ^[a] of the cell 4-1 is determined.

  For example, in the cell 4-4 of interest, the cells 4-3 and 4-5 are cells whose mutual interference / interference cannot be ignored. In this case, since these cells are all connected to the common control station 1-2, no information exchange is required on the wired transmission path 5-3 between the control station 1-1 and the control station 1-2. However, in reality, the calculation processing of the transmission signal in all cells is performed in parallel, so the information required in any cell is exchanged via the wired transmission path 5-3. Share.

  In the above embodiment, the case where there are a plurality of control stations has been described. However, if the processing of Expression (26) and Expression (27) described in the principle of the present invention is performed, the control station is not necessarily provided. There need not be multiple. To the last, the signal processing can be distributed to a plurality of control stations in a distributed manner if only necessary information exchange is performed according to the present invention in contrast to the conventional method in which signal processing is intensively performed by one control station. As a result, the amount of computation for each control station can be suppressed to a low level, and as a result, the circuit scale can be suppressed to a realizable level.

  In addition, the loop is repeated for the processing from step S25 to step S28 in FIG. 1 (c). However, in terms of hardware, an individual circuit may be mounted for each loop or the same circuit may be repeatedly used. Either one does not matter.

Further, in the above description, the partial channel matrix H ′ _{i, j} has been described as a matrix. However, if a simple scalar quantity is also understood as a 1 × 1 matrix, the partial channel matrix H ′ _{i, j} does not necessarily have to be a matrix. The expansion is also possible when the matrix H ′ _{i, j} is a scalar quantity. In this case, the reception weight W _{i, i} can also be regarded as a 1 × 1 matrix. In this case, the reception weight W ′ _{i, i} is calculated by the attenuation and phase of the signal generated on the propagation path. Can be regarded as a process of setting the reciprocal of the scalar quantity H ′ _{i, i} corresponding to the rotation quantity of W ′ _{i, i} . Even if the partial channel matrix H ′ _{i, j} is a vector, if it is understood as a matrix of 1 × m (m is an integer of 2 or more), the same extension is possible.

  Accordingly, all of the embodiments described above are merely illustrative of the present invention and are not intended to limit the present invention, and the present invention can be implemented in various other variations and modifications.

1-1 to 1-2 Control station
2-1 to 2-5 Remote base station
3-1 to 3-5 Terminal stations
4-1 to 4-5 Cells using the same frequency channel
5-1 to 5-3 Wired transmission line

Claims (2)

A wireless communication method in which one or more base stations are arranged in a plurality of cells, each base station is connected to a network via a wired line, and each base station communicates with a terminal station in the cell via a wireless line. There,
A first step of obtaining channel information between the base station and the terminal station in the same cell and between different cells;
A second step of managing the channel information acquired in the first step as a channel matrix H ′ _{i, j} between a base station in the i-th cell and a terminal station in the j-th cell;
A reception weight matrix W ′ _{i, i} between the base station and the terminal station in the same cell is obtained based on the channel matrix, and for the channel matrix,
G _{i, j} = W ′ _{i, i} H ′ _{i, j} (W ′ _{j, j} H ′ _{j, j} ) ⁻¹
A third step of obtaining a matrix G _{i, j}
When the reception information in the i-th cell is R _i , the 0th-order estimated signal S _i ^[0] = W ′ _{i, i} R _i in the _i -th cell is obtained by multiplication with the reception weight matrix W ′ _{i, i.} A fourth step of obtaining;
Fifth order estimation signal S _j ^[a] in the j-th cell with j ≠ i calculated in all or a part of the surrounding cells that communicate using the same frequency channel as the i-th cell is obtained. Steps,
The summation result of the estimated signal S _j ^[a] and the matrix G _{i, j} is calculated with respect to a predetermined j, and the result is subtracted from the estimated signal S _i ^[0] . A wireless communication method, wherein an estimated signal is generated by repeatedly performing a sixth step of calculating S _i ^{[a + 1]} for an a + 1-order estimated signal. The wireless communication device according to claim 1, wherein one or more base stations are arranged in each of a plurality of cells, each base station is connected to a network via a wired line, and each base station is connected to a terminal station in the cell via a wireless line. A wireless communication system characterized in that communication is performed by a communication method.

Images