JP2011101291A - Wireless base station, multi-user mimo system, and user selection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an efficient user selection method in a multi-user MIMO system. <P>SOLUTION: A base station is provided, which includes: a user classification unit which classifies users in a self-cell into boundary users who are located in the vicinity of the boundary between the adjacent cell adjacent to the self-cell and the self-cell, and internal users except the boundary users; a user selection unit which selects a first user whose received power becomes the maximum for every resource block, and selects a second user whose channel capacity with the first user, when the same resource block is utilized, is large for every resource block from among second users different from the first user from among the users classified by the user classification unit for each of the boundary users and the internal users; and a transmission unit which transmits a signal by using the resource blocks corresponding to the first and second users to the first and second users selected by a first user selection unit. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a base station, a multi-user MIMO system, and a user selection method. In particular, the present invention relates to a user scheduling method in a zero-forcing beamforming multi-user MIMO-OFDM system.

  As one of techniques for increasing the communication speed between wireless devices, a multiple input multiple output (MIMO) system is known. This method is literally based on signal input / output using a plurality of antennas. The feature of this method is that a plurality of transmission data can be transmitted at the same time and at the same frequency using a plurality of different antennas. Therefore, as the number of channels that can be transmitted simultaneously increases, the amount of information that can be transmitted per unit time can be increased by the increased number of channels. Further, this method has an advantage that the occupied frequency band does not increase when the communication speed is improved.

  However, since a plurality of modulated signals having carrier components of the same frequency are transmitted at the same time, a means for separating the interfering modulated signals on the receiving side is necessary. Therefore, on the receiving side, a channel matrix representing the transmission characteristics of the wireless transmission path is estimated, and the transmission signal corresponding to each substream is separated from the received signal based on the channel matrix. The channel matrix is estimated using a reference signal. As a method for detecting a signal using an estimated channel matrix, for example, a method using a MMSE (Minimum Mean Squared Error) detection method or an MLD (Maximum Likelihood Detection) detection method is known.

  Recently, multi-users improve the overall system throughput by transmitting signals to multiple radio terminals using multiple antennas installed in the radio base station, spatially multiplexing the signals, and receiving and separating the signals. A MIMO system is being used. Also, a distributed antenna radio access system (hereinafter referred to as a distributed antenna system) that disperses and arranges a plurality of antennas in one cell and implements radio communication between a radio base station and a radio terminal using the plurality of antennas. ) Is intended. For example, the following Non-Patent Document 1 describes a system configuration of a distributed antenna system to which a zero-forcing beamforming (hereinafter, ZFBF) multi-user MIMO transmission scheme is applied. Non-Patent Document 2 below describes an effective user selection method in the ZFBF multi-user MIMO transmission system.

CBPeel, BM Hochwald, and AL Swindlehurst ", A vector-perturbation technique for near-capacity multiantennamultiuser communication Part I: channel inversion and regulation," IEEETrans.Commun., Vol.53, no.1, pp.195-202, January2005 G. Dimic and N.D. Sidiropoulos, "On downlink beamforming with greedy userselection: performance analysis and a simple new algorithm", IEEE Trans.Signal Processing, vol.53, no.10, pp.3857-3868, October 2005

  In the technology described in Non-Patent Document 1 described above, in an environment where there are a plurality of users in the own cell managed by a certain radio base station, signals transmitted toward each user in the own cell are orthogonal to each other. Is applied with ZFBF. With such a configuration, it is possible to suppress interference of signals for each user in the downlink. However, the technique described in this document does not consider interference with other cells (hereinafter referred to as adjacent cells) adjacent to the own cell. However, in practice, when a user in the own cell is located in the vicinity of the adjacent cell, the distributed antenna system of the own cell is affected by interference due to the transmission signal of the adjacent cell. In particular, when a plurality of users are close to each other across a cell boundary, they influence each other and transmission characteristics are greatly deteriorated in any user.

  In addition, the technique described in Non-Patent Document 2 described above sequentially selects a user having the maximum received signal power and another user having a larger channel capacity based on the channel vector of each other user. It is. With such a configuration, a combination of users with high throughput is selected. However, users with a small channel capacity located at the cell boundary are hardly selected. As a result, fairness for the user cannot be maintained. In addition, this technique has a problem that the amount of calculation is large.

  Therefore, the present invention has been made in view of the above problems, and an object of the present invention is to provide a new and improved base capable of improving throughput while maintaining a user's fairness to some extent. It is to provide a station, a multi-user MIMO system, and a user selection method.

  In order to solve the above problem, according to an aspect of the present invention, a user in a local cell is divided into a boundary user located near a boundary between the adjacent cell adjacent to the local cell and the local cell, and the boundary user. A first boundary user that selects the first boundary user having the maximum received power for each resource block from among the internal user other than the user classification unit that classifies and the boundary user classified by the user classification unit A selection unit; a first internal user selection unit that selects a first internal user having the maximum reception power for each resource block from the internal users classified by the user classification unit; and the first boundary user The second boundary user having a large channel capacity with the first boundary user when the same resource block is used is selected from the second boundary users different from the second boundary user. A channel between the second boundary user selection unit selected for each resource block and the first internal user when the same resource block is used from the second internal users different from the first internal user A second internal user selection unit that selects a second internal user having a large capacity for each resource block, a first boundary user selected by the first boundary user selection unit, and a selection by the second boundary user selection unit A signal is transmitted to the second boundary user using the resource blocks corresponding to the first and second boundary users, and the first internal user selected by the first internal user selection unit and the A signal is sent to the second internal user selected by the second internal user selection unit using the resource blocks corresponding to the first and second internal users. And a transmission unit for transmitting the radio base station is provided.

  In this way, by classifying the users inside the cell and the cell boundary users, assigning resource blocks for each classification, and selecting spatially multiplexed users for each resource block, it is possible to perform spatial multiplexing while maintaining user fairness. Can improve throughput.

  In addition, the radio base station may further include a channel estimation unit that estimates a channel vector for each resource block. In this case, the first boundary user selection unit uses the reception power for each resource block calculated from the channel vector estimated by the channel estimation unit, and the first boundary where the reception power becomes maximum for each resource block. The first internal user selection unit selects a user, and uses the received power for each resource block calculated from the channel vector estimated by the channel estimation unit, so that the received power is maximized for each resource block. Select an internal user. By estimating the channel vector for each resource block and classifying users using the estimated channel vector, it becomes possible to suppress the influence of noise, so that cell boundary users and cell internal users can be more effectively Can be classified.

  The second boundary user selection unit includes a channel vector corresponding to the first boundary user estimated by the channel estimation unit, and a channel corresponding to the second boundary user estimated by the channel estimation unit. An inverse matrix of a channel matrix formed by the vector may be calculated, and a second boundary user having the maximum channel capacity may be selected based on a diagonal component of the inverse matrix. Further, the second internal user selection unit includes a channel vector corresponding to the first internal user estimated by the channel estimation unit, and a channel corresponding to the second internal user estimated by the channel estimation unit. An inverse matrix of a channel matrix formed by the vector may be calculated, and a second internal user having the maximum channel capacity may be selected based on the inverse matrix.

  Thus, when selecting a spatially multiplexed user, cell throughput can be improved by sequentially selecting users with a large channel capacity. In addition, since the selection of spatially multiplexed users is performed for each classification, the number of combinations of users is reduced by a smaller number of users to be subjected to the selection process than when spatially multiplexed users are selected for all users. As a result, the amount of calculation can be reduced by selecting a spatially multiplexed user for each classification. That is, by applying the above configuration, it is possible to reduce the amount of calculation required for the spatial multiplexing user selection process.

The second boundary user selection unit selects a second boundary user Uk having the k-th largest channel capacity for k = _{1 to} n ₁ (n ₁ <number of boundary users); A channel vector corresponding to the first boundary user, a channel vector corresponding to the second boundary user U _p (p = 1 to k), and the second boundary user U _p estimated by the channel estimation unit And an inverse matrix of a channel matrix formed by a channel vector corresponding to a second boundary user different from, and the second largest channel capacity is (k + 1) th based on the diagonal component of the inverse matrix a step of selecting a boundary user _{U (k + 1)} of the repeatedly performed, the second internal user selector, for l = 1~n _{2 (n} 2 <number of internal users), the channel A step of capacity to select the second internal users U _l large k-th, and the channel vector corresponding to the first internal users, corresponding to the second internal user U _{q (q} = 1~l) And an inverse matrix of a channel matrix formed by a channel vector corresponding to a second internal user different from the second internal user U _q estimated by the channel estimation unit, And a step of selecting the second internal user U _{(k + 1)} having the (k + 1) -th largest channel capacity based on the diagonal component of the inverse matrix. With such a configuration, it is possible to select a predetermined number of spatially multiplexed users in consideration of the channel state of the selected user while sequentially selecting spatially multiplexed users.

  The user classifying unit is calculated based on a desired signal power calculated based on a cell-specific downlink reference signal of the own cell and a signal power received by a resource block not used for signal transmission of the own cell. The boundary user and the internal user may be classified using an average SIR (Signal to Interference Ratio) of the entire resource block estimated by each user using interference power. When the average SIR of the entire resource block is obtained for each user, the cell boundary user and the cell internal user can be easily classified using the average SIR. Further, by using the average SIR, more practical classification can be performed in consideration of the channel environment as compared with the case of classifying users by physical distance from the base station.

  In addition, the user classifying unit sets a user whose average SIR exceeds a predetermined threshold as the internal user, sets a user other than the internal user as the boundary user, or sets a user in a descending order of the average SIR. A ratio of users may be set as the internal users, and users other than the internal users may be set as the boundary users. For example, if the percentage of users inside the cell is increased and the percentage of cell boundary users can be adjusted to be small, it is easy to reduce the partial band allocated to the cell boundary users while suppressing the influence on the cell boundary users. .

  In order to solve the above-described problem, according to another aspect of the present invention, an average SIR fed back from a user in the own cell is used to make a user in the own cell adjacent to the own cell. For each resource block, the user is classified into a boundary user located near the boundary between the cell and the own cell, an internal user other than the boundary user, and a boundary user classified by the user classification unit. A first boundary user selection unit that selects a first boundary user with the highest received power and a first user with the highest received power for each resource block among the internal users classified by the user classifying unit. The first boundary user when the same resource block is used from the first internal user selection unit for selecting an internal user and the second boundary user different from the first boundary user. A second boundary user selecting unit that selects, for each resource block, a second boundary user having a large channel capacity between the second internal user and the second internal user different from the first internal user, and the same resource block Is selected by a second internal user selection unit that selects, for each resource block, a second internal user having a large channel capacity with the first internal user, and the first boundary user selection unit. A signal is transmitted to the first boundary user and the second boundary user selected by the second boundary user selection unit using a resource block corresponding to the first and second boundary users, and The first internal user selected by one internal user selection unit and the second internal user selected by the second internal user selection unit A base station having a transmission unit that transmits a signal using a resource block corresponding to a second internal user, calculates a desired signal power based on a cell-specific downlink reference signal of the own cell, An SIR calculating unit that calculates interference power based on signal power received by a resource block that is not used for signal transmission, and calculates an average SIR of the entire resource block using the desired signal power and the interference power; There is provided a multi-user MIMO system including a user terminal having a feedback unit that returns the average SIR calculated by the SIR calculation unit to the base station.

  In this way, by classifying the users inside the cell and the cell boundary users, assigning resource blocks for each classification, and selecting spatially multiplexed users for each resource block, it is possible to perform spatial multiplexing while maintaining user fairness. Can improve throughput. Further, by classifying users based on the average SIR, it is possible to classify users more practically than when classifying users based on physical distance.

  In order to solve the above problem, according to another aspect of the present invention, a user in the own cell is a boundary user located in the vicinity of a boundary between the adjacent cell adjacent to the own cell and the own cell, The first boundary user that has the maximum reception power for each resource block is selected from among the user classification step that is classified into internal users other than the boundary user and the boundary users that are classified in the user classification step. A first boundary user selection step, a first internal user selection step of selecting a first internal user having the maximum received power for each resource block from the internal users classified in the user classification step, and the first The second boundary user having a large channel capacity with the first boundary user when the same resource block is used among the second boundary users different from the boundary user of the second boundary user. A second boundary user selection step of selecting the user for each resource block, and the first internal user when the same resource block is used from the second internal users different from the first internal user. There is provided a user selection method in a multi-user MIMO system, including a second internal user selection step of selecting a second internal user having a large channel capacity between each resource block.

  In this way, by classifying the users inside the cell and the cell boundary users, assigning resource blocks for each classification, and selecting spatially multiplexed users for each resource block, it is possible to perform spatial multiplexing while maintaining user fairness. Can improve throughput.

  As described above, according to the present invention, it is possible to improve throughput while maintaining user fairness to some extent.

It is explanatory drawing which shows the principle of a Round robin user selection method. It is explanatory drawing which shows the principle of Fractional Frequency Reuse (FFR). It is explanatory drawing which shows the principle of zero-forcing selection (ZFS) for Zero-forcing beamforming multiuser MIMO system. It is explanatory drawing which shows the function structural example of the base station which concerns on one Embodiment of this invention. It is explanatory drawing which shows an example of the allocation method of Cell-specific downlink RS which concerns on the embodiment. It is explanatory drawing which shows an example of the allocation method of Cell-specific downlink RS which concerns on the embodiment. It is explanatory drawing which shows the function structural example of the user terminal which concerns on the same embodiment. It is explanatory drawing which shows the classification method of Cell internal user / Cell boundary user which concerns on the embodiment. It is explanatory drawing which shows the allocation method to the Cell of the frequency band which concerns on the same embodiment. It is explanatory drawing which shows the RB selection method for every user which concerns on the embodiment. It is explanatory drawing which shows the selection method of the spatial multiplexing user which concerns on the embodiment. It is explanatory drawing which shows the example of a Cell structure and example of antenna arrangement concerning the example of application (application to a distributed antenna system) of the embodiment. It is explanatory drawing which shows the function structural example (part) of the base station which concerns on the application example. It is explanatory drawing which shows the function structural example (whole) of the base station which concerns on the application example. It is explanatory drawing which shows the simulation result which compared the average Cell throughput between the system which concerns on the same embodiment, and the conventional system. It is explanatory drawing which shows the simulation result which compared the average user throughput between the system which concerns on the same embodiment, and the conventional system.

  Exemplary embodiments of the present invention will be described below in detail with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, duplication description is abbreviate | omitted by attaching | subjecting the same code | symbol.

[About the flow of explanation]
Here, the flow of explanation regarding the embodiment of the present invention described below will be briefly described. First, a user selection method of the round robin method (hereinafter referred to as RR method) will be described with reference to FIG. Next, the concept of the fractional frequency reuse method (hereinafter referred to as FFR method) will be described with reference to FIG. Next, a user selection method (hereinafter referred to as ZFS; Zero-Forcing Selection) in the ZFBF multi-user MIMO system will be described with reference to FIG.

  Next, the functional configuration of the base station 10 according to an embodiment of the present invention will be described with reference to FIG. Among these, a method for assigning a cell-specific downlink reference signal (Cell-specific downlink RS) will be described with reference to FIGS. 5A and 5B. Further, the functional configuration of the user terminal 20 according to the present embodiment will be described with reference to FIG. Furthermore, with reference to FIG. 7, a method for classifying a cell internal user and a cell boundary user will be described. Then, a method of assigning cells to frequency bands will be described with reference to FIG. Further, a resource block (RB) allocation method for each user will be described with reference to FIG. Further, a method for selecting a user to be subjected to spatial multiplexing transmission will be described with reference to FIG.

  Next, application to a distributed antenna system will be described as an application example of the present embodiment. First, the cell configuration and the arrangement of distributed antennas in the distributed antenna system will be described with reference to FIG. Next, the functional configuration of the base station 30 according to this application example will be described with reference to FIGS. 12 and 13. Next, effects obtained by applying the technique according to the present embodiment will be described with reference to FIGS. 14 and 15.

[Organization of issues]
First, prior to a detailed description of a technique according to an embodiment of the present invention, problems to be solved by the embodiment will be briefly summarized.

(About RR method)
First, an RR user selection method in a MIMO-OFDM system will be described with reference to FIG. In the example of FIG. 1, it is assumed that 6 users are spatially multiplexed in each RB. In the RR scheme, every time the RB index (RB index; hereinafter referred to as RB index) increases by 1, the index of 6 users (for example, {1, 2, 3, 4, 5, 6}; hereinafter referred to as user index) is 1. Shift (for example, {2, 3, 4, 5, 6, 7}). Note that the user index may be cyclically replaced as the RB index increases by one.

  With this configuration, all users are equally allocated to RBs. Therefore, high fairness to the user is realized. However, in the case of the RR method, since the channel state with the user is not taken into consideration, the effect of improving the throughput cannot be expected.

(About FFR method)
Next, the FFR method will be described with reference to FIG. The FFR scheme is a technique devised to improve transmission quality of users located at cell edges or cell boundaries. Users located at cell edges or cell boundaries are greatly affected by inter-cell interference from neighboring cells. For this reason, the transmission quality of a user located at the cell edge or cell boundary is greatly deteriorated due to the influence of inter-cell interference. Then, the rate control reduces the coding rate and the modulation order of the error correction code, and the throughput is greatly reduced.

  Thus, in the FFR scheme, inter-cell interference is suppressed by imposing restrictions on users located at cell edges or cell boundaries so as to use partial bands predetermined for each cell. In the case of the FFR method, in principle, the frequency utilization efficiency is sacrificed, but the inter-cell interference is greatly reduced. As a result, the transmission quality is improved, and the effect of improving the frequency utilization efficiency is actually obtained.

  In the example of FIG. 2, FRF (Frequency Reuse Factor) = 1 is set for users inside the cell, and FRF = 3 is set for users located at the cell boundary. FRF corresponds to the number of frequency repetitions. In the example of FIG. 2, as shown in FIG. 2A, users located inside the cell (hereinafter referred to as cell internal user) and users located at the cell boundary (hereinafter referred to as cell boundary user) are classified, and (B ), The frequency band is divided and assigned.

  However, in the case of the FFR scheme, scheduling for each user is required after classifying the cell internal user and the cell boundary user. Moreover, since the partial band allocated to the cell boundary user is often set to be smaller than the partial band allocated to the cell internal user, there is a problem that fairness cannot be maintained for the cell boundary user.

(About ZFS method)
Next, the ZFS method will be described with reference to FIG. The ZFS scheme is a user selection technique devised to improve throughput in a ZFBF multi-user MIMO system. In FIG. 3, _NT represents the number of base station antennas, and N _U represents the number of users to be spatially multiplexed.

  As shown in FIG. 3, in the ZFS system, first, received signal power is calculated based on a channel vector between each user, and a user (hereinafter referred to as a first user) having the maximum received signal power is selected (hereinafter referred to as a first user). S21). Further, the channel vector of the first user is selected (S21). Next, one user is selected from the remaining users, and a channel matrix is generated using the channel vector of the user and the channel vector of the first user (S22).

  Next, the inverse matrix of the channel matrix is generated and calculated (S23), and the sum rate (sum of the channel capacity for the number of users) is calculated using the inverse of each diagonal component of the calculated inverse matrix as the channel gain after ZFBF. (S24). Next, the operations in steps S22 to S24 are performed for all the remaining users (S27), and the user (hereinafter referred to as the second user) having the maximum Sum rate is detected (S25, S26).

  Similarly, a channel matrix is generated using the channel vectors of the first and second users and the channel vector of one user other than the first and second users (S22), inverse matrix calculation (S23), and sum rate calculation. (S24), maximum sum rate user detection (S25, S26) is performed. A series of these operations are repeatedly executed to select a combination of NU users that increase the channel capacity (S28).

  By performing the operations in steps S21 to S28, a combination of users for increasing the channel capacity is selected, so that an effect of improving the throughput can be expected. However, since there are users who are hardly selected, the fairness of the users cannot be maintained.

<Embodiment>
In view of the above problems, the present inventors have devised a method capable of improving throughput while maintaining user fairness. Hereinafter, an embodiment of the present invention related to such a method will be described.

[Configuration of base station 10]
First, the functional configuration of the base station 10 according to the present embodiment will be described with reference to FIG. FIG. 4 is an explanatory diagram illustrating a functional configuration example of the base station 10 according to the present embodiment. The base station 10 is an example of a radio base station included in the multiuser MIMO system. In the following description, the cell arrangement shown in FIG. 5A and the RB configuration (downlink reference signal allocation configuration) shown in FIG. 5B are assumed as an example.

  As shown in FIG. 4, the base station 10 mainly includes a user classifying unit 102 (FFR user classifying unit), an RB selecting unit 104 (User fairness RB selecting unit), and a user selecting unit 106 (for High capacity). a user selection unit), spatial multiplexing units 108 and 110 (per-RB ZFBF MU-MIMO NU user spatial multiplexing unit), and an OFDM unit 112 (antenna per OFDM unit). The base station 10 has NT antennas.

(User classification unit 102)
First, the user classification unit 102 will be described. The user classifying unit 102 is means for reducing the influence of inter-cell interference that is received by cell boundary users. Based on the FFR scheme, the user classifying unit 102 classifies the users in the own cell into cell internal users (inner-cell user sets) and cell boundary users (Cell-edge user sets). Note that an average SIR (Signal to Interference Ratio; the average SIR returned from the kth user is expressed as SIR _k ) fed back from the user terminal 20 (described later) of each user is input to the user classifying unit 102. Is done.

  As shown in FIG. 7, the user classifying unit 102 ranks the users in descending order of the average SIR returned from the user terminal 20 (S31). Further, the user classifying unit 102 classifies users with high ranks as in-cell users (S32), and classifies other users as cell boundary users (S33). However, the classification may use a predetermined threshold value or a predetermined distribution ratio.

  For example, users whose average SIR exceeds a predetermined threshold are classified as in-cell users, and other users are classified as cell boundary users. In addition, users within a cell may be set according to a predetermined distribution ratio, a user whose average SIR is as high as that number may be classified as a cell internal user, and other users may be classified as cell boundary users. Note that in FIG. 7B, it seems that the cell internal user and the cell boundary user are classified according to the distance, but it is not necessarily classified according to the geographical distance in practice. I want to be.

(RB selection unit 104)
Next, the RB selection unit 104 will be described. The RB selection unit 104 is a means for maintaining user fairness. When a user is classified by the user classification unit 102, the RB selection unit 104 selects an RB for each user of each classification. Note that information on the FFR cell-specific partial band is input to the RB selection unit 104. For example, the information of the band f and FRF = 1 shown in FIG. 8 is input as the partial band information related to the user inside the cell, and the bands f ₁ , f ₂ and f shown in FIG. _3. Information of FRF = 3 is input.

  As shown in FIG. 9A, the RB selection unit 104 calculates the received signal power for each RB, using the estimated value of the channel vector estimated from the sounding reference signal by each user for the users in the cell (S41). ). Next, as shown in FIG. 9B, the RB selection unit 104 compares the maximum value of the received signal power in each RB band (hereinafter referred to as the maximum received signal power) for each user, and the maximum received signal power is large. The users are ranked in order (S42).

  Next, the RB selection unit 104 compares the maximum received signal power of each user among the RBs in descending order from the highest ranked user, and selects the RB having the maximum maximum received signal power (S43). However, when the selected RB is selected as the RB of another user, the RB selection unit 104 selects the RB having the second largest maximum received signal power (S44). Note that the RB selection unit 104 may be configured to select an RB redundantly when the selected RB is selected as an RB of another user.

  Similarly, as shown in FIG. 9A, the RB selection unit 104 calculates the received signal power for each RB for each cell boundary user using the channel vector estimation value estimated from the sounding reference signal by each user. (S45). Next, as shown in FIG. 9B, the RB selection unit 104 compares the maximum received signal power in each RB band for each user, and ranks the users in descending order of the maximum received signal power (S46).

  Next, the RB selection unit 104 compares the maximum received signal power of each user among the RBs in descending order from the highest ranked user, and selects the RB having the maximum maximum received signal power (S47). However, when the selected RB is selected as the RB of another user, the RB selection unit 104 selects the RB having the second largest maximum received signal power (S48). Note that the RB selection unit 104 may be configured to select an RB redundantly when the selected RB is selected as an RB of another user.

  By the operations in steps S41 to S44 and S45 to S48, one user is assigned to each RB. Thus, the received signal power is calculated from the channel vector for each RB estimated in the TDD (Time Division Duplex) uplink, and the RB for which the maximum received signal power is obtained for each user is calculated based on the calculation result. By selecting, it contributes to the improvement of the throughput. Moreover, since all the target users are selected, the user's fairness can be maintained.

(User selection unit 106)
Next, the user selection unit 106 will be described. The user selection unit 106 is a means for improving throughput. Based on the channel vector of the user selected by the RB selection unit 104 for each RB, the user selection unit 106 selects a user having a channel vector orthogonal to the channel vector and having a large channel gain.

  As shown in FIG. 10, the RB selection unit 104 selects the user selected for each RB (first user) and the channel vector of the first user (S51). First, the user selection unit 106 selects a channel vector of another user as a candidate (S57), and generates a channel matrix from the channel vector of the first user and the candidate channel vector (S52). At this stage, a two-user channel matrix is generated. Next, the user selection unit 106 calculates an inverse matrix of the generated channel matrix (S53). Next, the user selection unit 106 calculates Sum rate from the inverse of the calculated diagonal component of the inverse matrix (S54).

  Next, the user selection unit 106 calculates Sum rate by repeating the operations of Steps S52 to S54 described above for channel vector candidates of other users different from the first user (S57). And the user selection part 106 detects the maximum value of the calculated Sum rate, and selects the user corresponding to the maximum value to a 2nd user (S55, S56).

Next, the user selection unit 106 selects a channel vector of another user different from the first and second users as a candidate (S57, S58), and uses the channel vector of the first and second users and the candidate channel vector. A channel matrix is generated (S52), an inverse matrix calculation (S53), a sum rate calculation (S54), and a user selection that maximizes the sum rate (S55, S56). Similarly, the user selection unit 106, by executing the above steps S52~S58 sequentially selecting a combination of _{N U} users.

  By such an operation, a combination of users whose throughput is expected to be maximized when spatially multiplexed is selected on the basis of the channel vector of the user selected by the RB selection unit 104 for each RB. By using this combination to transmit spatially multiplexed signals, an effect of improving throughput is expected. In addition, since users are classified into cell internal users and cell boundary users in advance and ZFS is performed for each classification, the number of combinations of users is greatly reduced by the small number of target users of ZFS. On the other hand, the total calculation amount is greatly reduced as compared with the case of applying ZFS.

(Spatial multiplexing units 108 and 110, OFDM unit 112)
Next, the spatial multiplexing units 108 and 110 and the OFDM unit 112 will be described. When the user selection unit 106 selects a combination of users for each classification and for each RB, the spatial multiplexing units 108 and 110 perform channel coding and modulation processing for each RB on the data to be transmitted to each user. Furthermore, transmission beam forming is performed and spatial multiplexing is performed. Next, the OFDM unit 112 performs orthogonal frequency division multiplexing on the resource elements related to all RBs using IFFT (Inverse Fast Fourier Transform) and transmits the resource elements.

  Heretofore, the configuration of the base station 10 according to the present embodiment has been described. As described above, the base station 10 can realize a user selection method including three elements: user classification based on average SIR, RB selection for each user in each classification, and spatial multiplexing user selection for each RB. Since user classification is performed and a combination of RBs and spatially multiplexed users is selected for each classification, it is possible to improve throughput while maintaining user fairness.

  By the way, when the distribution ratio is used when classifying the cell internal user and the cell boundary user, the number of users assigned to the partial band for the cell boundary user is reduced by reducing the number of users classified as the cell boundary user. Therefore, even if the partial band for the cell boundary user is set small, the low fairness of the cell boundary user can be compensated. Further, by increasing the number of users of the cell boundary users that are spatially multiplexed, the low fairness of the cell boundary users can be compensated even if the partial band for the cell boundary users is set small. As a result, it is possible to reduce the inter-cell interference received by the cell boundary users, and obtain the effect of improving the throughput by spatial multiplexing while maintaining the fairness of resource allocation to each user.

[Supplement: Reference signal assignment method]
Here, a reference signal allocation method will be described with reference to FIGS. 5A and 5B. Note that the cell arrangement illustrated in FIG. 5A (note that the components having the same concentration hatching in FIGS. 5A, 5B, and 6 have a corresponding relationship with each other) is assumed. In addition, it is assumed that communication by TDD is performed. Based on this premise, the reference signal can be transmitted using the same frequency in the uplink and downlink. Also, in a TDD downlink, a downlink reference signal common to all base station antennas is used. With such a configuration, even if the number of base station antennas increases, the frame efficiency does not decrease.

  As shown in FIG. 5B, in the present embodiment, a restriction is imposed so that cell-specific downlink reference signals (Cell-specific downlink RS) are allocated to different positions in time / frequency space (resources) in adjacent cells. Furthermore, a restriction is imposed so that no signal is assigned to a position in the time / frequency space where a cell-specific downlink reference signal is assigned in an adjacent cell. By imposing such a restriction, nothing is transmitted to the position where the cell-specific downlink reference signal is allocated in the adjacent cell. This eliminates the influence of inter-cell interference from neighboring cells. As will be described later, the average SIR can be easily estimated by applying such a cell-specific downlink reference signal allocation method.

[Configuration of User Terminal 20]
Next, the functional configuration of the user terminal 20 according to the present embodiment will be described with reference to FIG. FIG. 6 is an explanatory diagram illustrating a functional configuration example of the user terminal 20 according to the present embodiment. The functional configuration of the user terminal 20 relates to an average SIR estimation method used for the above-described user classification.

  As shown in FIG. 6, the user terminal 20 mainly includes a GI removal & FFT unit 202, a channel vector estimation unit 204, an average power estimation unit 206, an interference wave average power estimation unit 208, and an average SIR estimation unit 210. And the quantization unit 212.

First, the user terminal 20 receives the OFDM signal transmitted by the base station 10 using a plurality of terminal antennas (# 1 to #n _R ). This OFDM signal is input to the GI removal & FFT unit 202. When the guard interval is added to the input OFDM signal, the GI removal & FFT unit 202 removes the guard interval. Further, the GI removal & FFT unit 202 performs fast Fourier transform (FFT) on the OFDM signal to convert it to a subcarrier signal.

  The output of the conversion process in the GI removal & FFT unit 202 is input to the channel vector estimation unit 204. When a subcarrier signal is input, channel vector estimation section 204 estimates a channel vector using a cell-specific downlink reference signal multiplexed on the input subcarrier signal. The channel vector estimated using the cell-specific downlink reference signal of the own cell (for example, Cell # 1) is input to the desired wave average power estimation unit of average power estimation unit 206. On the other hand, channel vectors estimated using the cell-specific downlink reference signals of other cells (for example, Cell # 2, # 3, # 4) are input to the interference wave average power estimation unit of the corresponding average power estimation unit 206, respectively. Is done.

  The desired wave average power estimator estimates the received power of the desired wave based on the channel vector for each subcarrier input from the channel vector estimator 204. Similarly, each interference wave average power estimation unit estimates the reception power of the interference wave based on the channel vector for each subcarrier input from channel vector estimation unit 204.

  As described above, the cell-specific downlink reference signal of the adjacent cell is assigned to the RB to which nothing is transmitted in the own cell. For example, cell-specific downlink reference signals of Cell # 2, # 3, and # 4 are assigned to RBs to which nothing is transmitted in Cell # 1. In this case, the received signal power in the RB to which the cell-specific downlink reference signal of Cell # 1 is assigned is the sum of the signal power of the desired wave and the power of the interference signal received from non-adjacent Cells # 14 to # 19. It becomes. Also, the received signal power in the RB to which the cell-specific downlink reference signal of Cell # 1 is not allocated is the power of the interference signal received from adjacent Cell # 3, # 6, # 8, # 11, Cell # 4, This is the sum of the interference signal power received from # 7, # 9, and # 12.

  The desired wave average power estimation unit approximates the signal power received by the RB to which the cell-specific downlink reference signal of the own cell is assigned, and inputs the signal power to the average SIR estimation unit 210. Each interference wave average power estimation unit inputs the sum of received power received by each RB to which the cell-specific downlink reference signal of the own cell is not assigned to interference wave average power estimation unit 208. Interference wave average power estimation section 208 receives interference signal power (hereinafter referred to as other cell interference signal power) obtained from another cell, which is an average value obtained by averaging the sum of received power input from each interference wave average power estimation section over all RBs. To the average SIR estimation unit 210. The average SIR estimation unit 210 calculates an average SIR from the desired signal power and the other cell interference signal power, and inputs the average SIR to the quantization unit 212. The quantization unit 212 quantizes the average SIR for feedback and returns it to the base station 10.

  Heretofore, the functional configuration of the user terminal 20 according to the present embodiment has been described. As described above, in this embodiment, since the cell-specific downlink reference signal allocation method is devised so that the cell-specific downlink reference signals of neighboring cells do not interfere with each other, the average SIR can be easily estimated. As a result, it becomes possible to easily classify the cell internal user and the cell boundary user in the base station 10. Note that since cell-specific downlink reference signals are not subjected to transmission beamforming, the channel characteristics are different from those of data transmission channels subjected to transmission beamforming. However, by using the technology according to the present embodiment, it is possible to detect a user who is potentially susceptible to inter-cell interference. The user detected in this way is assigned to a partial frequency band of FRF ≧ 2, and the influence of inter-cell interference is reduced.

[Application example: Application to distributed antenna system]
Next, an application example in which the technique of the present embodiment is applied to a distributed antenna system will be described with reference to FIGS. FIG. 11 is an explanatory diagram showing a cell configuration and an antenna arrangement example in the distributed antenna system. 12 and 13 are explanatory diagrams illustrating an example of a functional configuration of the base station 30 according to this application example.

  First, referring to FIG. As shown in FIG. 11, in the distributed antenna system assumed in this application example, a large number of base station antennas are evenly arranged in a cell. Each cell is divided into a cell internal region and a cell boundary region. However, the classification between the inside of the cell and the cell boundary is not necessarily based on the geographical distance, and is classified based on the average SIR described above, for example. In the example of FIG. 11, the same hatching is used for the same frequency band, and different hatching parts are used for different frequency bands. In addition, it is assumed that the partial bands allocated to the cell boundary users are set so as not to overlap with each other by repeating the 3-cell frequency.

  Reference is now made to FIG. FIG. 12 illustrates in detail a functional configuration related to user selection (a user selection unit 312 described later) among the functional configurations of the base station 30. As illustrated in FIG. 12, the user selection unit 312 includes a user classification unit 302 (FFR user classification unit), an RB selection unit 304 (User fairness RB selection unit), and a user selection unit 306 (High-capacity user. Selection section). The basic configuration is the same as that of the base station 10 described above.

  The user classifying unit 302 performs ranking using the average SIR fed back from each user, classifies users with higher rankings as in-cell users, and classifies users with lower rankings as cell boundary users. In addition, the user classification unit 302 notifies the RB selection unit 304 of the ranking information of the users inside the cell and the ranking information of the cell boundary users. The RB selection unit 304 calculates received signal power for each RB, and selects a user to be assigned to each RB based on the calculated received signal power. At this time, by using the ranking information notified from the user classifying unit 302, the RB selection process can be simplified by selecting RBs in order from the top ranking users.

When one user is selected for each RB by the RB selection unit 304, the user selection unit 306 selects a combination of users spatially multiplexed in each RB. The user selection unit 306, based on the channel vector of the user selected by the RB selection unit 304, Sum Rate is sequentially selects the user having the maximum using the channel vectors of other users, the channel capacity is large N _U Select a combination of users. The combination of N _U users (N _U user indices) selected by the user selection unit 306 is output to the subsequent processing unit for each classification and for each RB.

  Reference is now made to FIG. As illustrated in FIG. 13, the base station 30 includes a user selection unit 312 (User selection unit), an AMC unit 314, a user-specific reference signal multiplexing unit 316 (User-specific RS multiplexing unit), a ZFBF unit 318, an antenna, A common cell specific downlink reference signal multiplexing unit 320, an OFDM unit 322, an FFT unit 324, and a channel vector estimation unit 326 (K user channel vector estimation unit) are configured. Here, a ZFBF multi-user MIMO-OFDM system is assumed.

First, in antenna common cell specific downlink reference signal multiplexing section 320, a downlink reference signal that is common among _NT base station antennas is generated. Then, antenna common cell specific downlink reference signal multiplexing section 320 multiplexes the downlink reference signal with the transmission data to generate a transmission frame. The transmission frame generated by antenna common cell specific downlink reference signal multiplexing section 320 is input to OFDM section 322 corresponding to each base station antenna. The OFDM unit 322 performs serial-parallel conversion on the transmission frame, converts the parallelized bit string into a plurality of subcarrier signals orthogonal to each other, and then performs IFFT on the subcarrier signal to generate an OFDM signal. The OFDM signals generated by the OFDM unit 322 are transmitted from the corresponding base station antennas.

  The OFDM signal transmitted from the base station antenna is received by each user terminal having a plurality of terminal antennas. The OFDM signal received by a plurality of terminal antennas is subjected to FFT in each user terminal, subjected to FFT, and converted to a subcarrier signal. Further, each user terminal estimates a channel vector based on the downlink reference signal multiplexed on the subcarrier signal. When the channel vector is estimated, a complex conjugate of the channel vector is calculated, and power normalization is performed to generate a beamforming weight vector (hereinafter referred to as BF weight vector) for the uplink reference signal.

  In each user terminal, a Zadoff-Chu sequence is generated as an uplink reference signal. The Zadoff-Chu sequence is an example of a sequence having excellent autocorrelation characteristics. A plurality of sequences obtained by cyclically shifting a Zadoff-Chu sequence is an example of a sequence having excellent cross-correlation characteristics. When each user terminal generates a Zadoff-Chu sequence as an uplink reference signal, if there is another cell that uses the same subcarrier as its own cell, a sequence having a sequence index different from that of the other cell is used as the uplink reference signal. Use. Then, a sequence having a different shift index for each user in the same cell is generated as an uplink reference signal.

  Next, the BF weight vector is multiplied with the uplink reference signal generated in this way. Then, IFFT is applied to the output of the product operation, a guard interval is added, and a distributed FDMA signal is generated. The generated distributed FDMA signal (uplink reference signal) is transmitted to the base station 30 via the TDD uplink via a plurality of terminal antennas.

  The uplink reference signal transmitted from each user terminal is received via a plurality of base station antennas that the base station 30 has. The uplink reference signal received via the base station antenna is input to the FFT unit 324, subjected to FFT, and then input to the channel vector estimation unit 326. Channel vector estimation section 326 separates uplink reference signals received from a plurality of user terminals for each user, and performs channel estimation for each resource block along the frequency direction. The channel vector obtained by channel estimation by the channel vector estimation unit 326 is input to the user selection unit 312. Also, the user selection unit 312 receives information on average SIR returned from each user terminal.

When the channel vector is inputted, the user selection unit 312, using the input channel vector, data transmission throughput after the downlink transmission beamforming so that a maximum, selects the N _U user. At this time, the user selection unit 312 maintains the fairness of the user with the configuration illustrated in FIG. 12 based on the channel vector input by the channel vector estimation unit 326 and the average SIR information fed back from each user terminal. Meanwhile, a combination of _NU users that improves the throughput is selected. The combination of users selected by the user selection unit 312 is input to the ZFBF unit 318.

ZFBF unit 318 calculates a transmission beam forming weighting matrix for the combination of _{N U} users selected by the user selecting unit 312. Further, the ZFBF unit 318 receives a downlink reference signal unique to each user generated by the user-specific reference signal multiplexing unit 316. Unlike the downlink reference signal generated by the antenna common cell specific downlink reference signal multiplexing unit 320, the downlink reference signal generated by the user specific reference signal multiplexing unit 316 is unique to each user selected by the user selection unit 312. It is a downlink reference signal. The ZFBF unit 318 performs transmission beam forming on the downlink reference signal using the previously calculated transmission beam forming matrix, and transmits from the plurality of base station antennas via the OFDM unit 322.

  The OFDM signal transmitted from the base station 30 is received by a plurality of terminal antennas of each user terminal, the guard interval is removed, and the OFDM signal is converted into a subcarrier signal. Then, the channel matrix after beamforming is estimated from the user-specific downlink reference signal transmitted toward the user and the user-specific downlink reference signal transmitted toward another user. The channel matrix estimated here is used for calculation of an MMSE reception beamforming weight vector (hereinafter referred to as a reception BF weight vector) for removing inter-user interference due to insufficient estimation accuracy of channel estimation values and channel time fluctuation.

  When the reception BF weight vector is calculated using the channel matrix, SINR (Signal to Noise Interference Ratio) after reception beamforming is estimated from the reception BF weight vector. The estimated SINR is used for selecting a Modulation and Coding Set (MCS) that can be decoded without error and that can achieve the maximum transmission rate. The MCS information (MCS index) selected using the estimated SINR value is fed back to the base station 30.

  The MCS index fed back from each user terminal is input to the AMC unit 314. When the MCS index is input, the AMC unit 314 performs error correction coding and modulation mapping on the transmission data based on the input MCS index, and inputs the transmission data to the ZFBF unit 318. AMC is an abbreviation for Adaptive Modulation and Coding. As the error correction code, for example, a Turbo code is used. The ZFBF unit 318 performs transmission beamforming on the transmission data that has been subjected to error correction coding and modulation mapping by the AMC unit 314 and transmits the transmission data via the OFDM unit 322. At this time, OFDM section 322 transmits all RB resource elements by orthogonal frequency division multiplexing using IFFT.

  The application example of the present embodiment has been described above. By applying the technology according to the present embodiment including the above application example, when assigning users to each RB related to frequency multiplexing and each stream related to spatial multiplexing in the ZFBF multi-user MIMO-OFDM system, It is possible to satisfy the three requirements of inter-cell interference reduction for users located, fairness of resource allocation to users, and improvement of throughput by spatial multiplexing. As a result, it contributes to the effective operation of the system.

[effect]
Finally, effects obtained by applying the configuration according to the present embodiment will be described with reference to FIGS. 14 and 15. FIG. 14 shows a comparison between the average cell throughput characteristics obtained by applying the ZFS scheme and the RR scheme and the average cell throughput characteristics obtained by applying the user selection scheme (hereinafter referred to as the PR scheme) according to the present embodiment. The simulation results are shown. FIG. 15 shows a simulation result regarding a comparison between an average user throughput characteristic obtained by applying the ZFS method and the RR method and an average user throughput characteristic obtained by applying the PR method. The simulation conditions are 24 users, 100 RBs (76 RBs with FRF = 1, RBs with FRF = 3 = 8), and 6 spatial multiplexing.

  As shown in FIG. 14, the average cell throughput characteristics are good in the order of the ZFS method, the PR method, and the RR method. In the case of the RR scheme, since the channel state is not considered, the average cell throughput is naturally low. On the other hand, in the case of the ZFS system, a combination of users having a large channel capacity is selected with the highest priority, so that users with poor channel conditions are hardly selected, and the average cell throughput becomes very high. However, as shown in FIG. 15, in the case of the ZFS method, the difference in average user throughput is significant.

  The horizontal axis of the graph shown in FIG. 15 indicates the user index. Also, user indexes are allocated in order from the user with the highest user throughput. Therefore, the user with the user index 1 is the user with the highest user throughput, and the user with the user index 24 is the user with the lowest user throughput. In other words, by comparing the average user throughput between the user of the user index 1 (hereinafter referred to as user # 1) and the user of the user index 24 (hereinafter referred to as user # 24), the degree of fairness of the user is evaluated. Can do.

  As described above, in the case of the ZFS method, a user with a high channel capacity is preferentially selected, and thus a high cell throughput is obtained. However, the difference between the user throughput of user # 1 and the user throughput of user # 24 is 10 It is about 10,000 times. On the other hand, in the case of the RR method, although the cell throughput is low, the difference between the user throughput of the user # 1 and the user throughput of the user # 24 is suppressed to about 10 times. On the other hand, in the case of the PR method, the difference between the user throughput of the user # 1 and the user throughput of the user # 24 is suppressed to about 30 times. Although the average cell throughput is not as good as that of the ZFS method, very good characteristics are obtained as compared with the RR method.

  Thus, by applying the technique according to the present embodiment, it is possible to improve throughput while maintaining user fairness.

  As mentioned above, although preferred embodiment of this invention was described referring an accompanying drawing, it cannot be overemphasized that this invention is not limited to the example which concerns. It will be apparent to those skilled in the art that various changes and modifications can be made within the scope of the claims, and these are naturally within the technical scope of the present invention. Understood.

DESCRIPTION OF SYMBOLS 10 Base station 102 User classification part 104 RB selection part 106 User selection part 108,110 Spatial multiplexing part 112 OFDM part 20 User terminal 202 GI removal & FFT part 204 Channel vector estimation part 206 Average power estimation part 208 Interference wave average power estimation part 210 Average SIR estimation unit 212 Quantization unit 30 Base station 302 User classification unit 304 RB selection unit 306, 312 User selection unit 314 AMC unit 316 User specific reference signal multiplexing unit 318 ZFBF unit 320 Antenna common cell specific downlink reference signal multiplexing unit 322 OFDM Unit 324 FFT unit 326 channel vector estimation unit

Claims (8)

A user classifying unit for classifying users in the own cell into border users located near the border between the neighboring cell adjacent to the own cell and the own cell, and internal users other than the border user;
A first boundary user selection unit that selects a first boundary user having the maximum received power for each resource block from the boundary users classified by the user classification unit;
A first internal user selection unit for selecting a first internal user having the maximum received power for each resource block from the internal users classified by the user classification unit;
For each resource block, a second boundary user having a large channel capacity with the first boundary user when the same resource block is used among the second boundary users different from the first boundary user is assigned to each resource block. A second boundary user selection unit to select,
For each resource block, a second internal user having a large channel capacity with the first internal user when the same resource block is used among the second internal users different from the first internal user. A second internal user selection unit to select
For the first boundary user selected by the first boundary user selection unit and the second boundary user selected by the second boundary user selection unit, resource blocks corresponding to the first and second boundary users are The first and second internal users selected by the first internal user selection unit and the second internal user selection unit selected by the first internal user selection unit are transmitted to the first and second internal users. A transmission unit that transmits a signal using a resource block corresponding to an internal user of
A radio base station comprising: A channel estimation unit for estimating a channel vector for each resource block;
The first boundary user selection unit selects a first boundary user having the maximum reception power for each resource block using reception power for each resource block calculated from the channel vector estimated by the channel estimation unit. And
The first internal user selection unit selects a first internal user having the maximum reception power for each resource block using reception power for each resource block calculated from the channel vector estimated by the channel estimation unit. The radio base station according to claim 1, wherein: The second boundary user selection unit includes:
The inverse of the channel matrix formed by the channel vector corresponding to the first boundary user estimated by the channel estimation unit and the channel vector corresponding to the second boundary user estimated by the channel estimation unit Calculate a matrix and select a second boundary user with the largest channel capacity based on the diagonal component of the inverse matrix;
The second internal user selection unit is
The inverse of the channel matrix formed by the channel vector corresponding to the first internal user estimated by the channel estimation unit and the channel vector corresponding to the second internal user estimated by the channel estimation unit The radio base station according to claim 2, wherein a matrix is calculated, and a second internal user having a maximum channel capacity is selected based on the inverse matrix. The second boundary user selection unit may perform k = _{1 to} n ₁ (n ₁ <number of boundary users).
Selecting a second boundary user U _k with the k-th largest channel capacity;
A channel vector corresponding to the first boundary user, a channel vector corresponding to the second boundary user U _p (p = 1 to k), and the second boundary user U estimated by the channel estimation unit. An inverse matrix of a channel matrix formed by a channel vector corresponding to a second boundary user different from _p is calculated, and the channel capacity is (k + 1) th largest based on the diagonal component of the inverse matrix Selecting two boundary users U _{(k + 1)} ;
Repeatedly
The second internal user selection unit is configured such that l = 1 to n ₂ (n ₂ <number of internal users).
A step of the channel capacity to select the second internal users U _l large k-th,
A channel vector corresponding to the first internal user, a channel vector corresponding to the second internal user U _q (q = 1 to l), and the second internal user U estimated by the channel estimation unit. An inverse matrix of a channel matrix formed by a channel vector corresponding to a second internal user different from _q is calculated, and the channel capacity having the (k + 1) th largest channel capacity is calculated based on the diagonal component of the inverse matrix. Selecting two internal users U _{(k + 1)} ;
The radio base station according to claim 3, wherein the radio base station is repeatedly executed. The user classifying unit
Using the desired signal power calculated based on the cell-specific downlink reference signal of the own cell and the interference power calculated based on the signal power received by the resource block not used for signal transmission of the own cell The said boundary user and the said internal user are classified using the average SIR (Signal to Interference Ratio) of the whole resource block estimated by each user, The any one of Claims 1-4 characterized by the above-mentioned. The radio base station described. The user classifying unit
A user whose average SIR exceeds a predetermined threshold is set as the internal user and a user other than the internal user is set as the boundary user, or a predetermined percentage of users in the descending order of the average SIR is set as the internal user. The radio base station according to claim 5, wherein a user other than the internal user is set as the boundary user. Using the average SIR fed back from the users in the own cell, the users in the own cell are classified as border users located near the border between the neighboring cell adjacent to the own cell and the own cell, and other than the border user. An internal user, a user classifying unit for classifying into,
A first boundary user selection unit that selects a first boundary user having the maximum received power for each resource block from the boundary users classified by the user classification unit;
A first internal user selection unit for selecting a first internal user having the maximum received power for each resource block from the internal users classified by the user classification unit;
For each resource block, a second boundary user having a large channel capacity with the first boundary user when the same resource block is used among the second boundary users different from the first boundary user is assigned to each resource block. A second boundary user selection unit to select,
For each resource block, a second internal user having a large channel capacity with the first internal user when the same resource block is used among the second internal users different from the first internal user. A second internal user selection unit to select
For the first boundary user selected by the first boundary user selection unit and the second boundary user selected by the second boundary user selection unit, resource blocks corresponding to the first and second boundary users are The first and second internal users selected by the first internal user selection unit and the second internal user selection unit selected by the first internal user selection unit are transmitted to the first and second internal users. A transmission unit that transmits a signal using a resource block corresponding to an internal user of
Having a base station,
Calculate desired signal power based on a cell-specific downlink reference signal of the own cell, calculate interference power based on signal power received by a resource block not used for signal transmission of the own cell, and calculate the desired signal power and An SIR calculation unit that calculates an average SIR (Signal to Interference Ratio) of the entire resource block using the interference power;
A feedback unit that returns the average SIR calculated by the SIR calculation unit to the base station;
A user terminal having:
A multi-user MIMO system characterized by comprising: A user classification step of classifying users in the own cell into border users located near the border between the neighboring cell adjacent to the own cell and the own cell, and internal users other than the border user;
A first boundary user selection step of selecting a first boundary user having the maximum received power for each resource block from the boundary users classified in the user classification step;
A first internal user selection step of selecting a first internal user having the maximum received power for each resource block from the internal users classified in the user classification step;
For each resource block, a second boundary user having a large channel capacity with the first boundary user when the same resource block is used among the second boundary users different from the first boundary user is assigned to each resource block. A second boundary user selection step to select,
For each resource block, a second internal user having a large channel capacity with the first internal user when the same resource block is used among the second internal users different from the first internal user. A second internal user selection step to select
A user selection method in a multi-user MIMO system, comprising:

Images