JP2010136453A - Wireless communication device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To derive a suitable weight in a communication system such as mobile WiMAX. <P>SOLUTION: The wireless communication device for performing wireless communication using the OFDMA system, includes an SMI weight derivation part 14 for deriving an SMI weight, and a weight integration part 15 for integrating the SMI weight to a received signal vector. The SMI weight derivation part 14 derives the SMI weight for every user allocation region (minimum unit region) by using a plurality of received subcarrier signal vectors contained in the user allocation region, preferably, in the minimum unit region of the user allocation region in the OFDMA system as sample values. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

The present invention relates to radio communications equipment.

  In recent years, WiMAX (Worldwide Interoperability for Microwave Access, IEEE802.16), which is one of the next generation wireless communication standards, has attracted attention. WiMAX is a standard that expands the communication speed and mobility of a wireless local area network (LAN) (see Non-Patent Document 1).

WiMAX employs an OFDMA (Orthogonal Frequency Division Multiple Access) system.
The OFDMA system is an extension of the OFDM (Orthogonal Frequency Division Multiplexing) system, which multiplexes data by dispersing data on a plurality of subcarriers (frequency) and multiplexing the frequency to multiple users.

  Further, the WiMAX communication system has a cell configuration, and each cell has a single base station (BS) installed on the roof of a building. A mobile station (MS: Mobile Station) communicates with a base station that desires communication.

  In this WiMAX uplink, a base station receives a signal from a mobile station, but a signal from a mobile station communicating with a base station in another cell is an interference wave (a wave from a mobile station not wishing to communicate). It is assumed that The presence of such interference waves degrades the quality of the received signal at the base station. For this reason, in order to improve the quality of the received signal, it is necessary to remove the interference wave.

One method for removing interference waves is an adaptive array. In an adaptive array, a signal is received by a plurality of antenna elements, and a desired signal is output by controlling and synthesizing the amplitude and phase of the signal received by each antenna element.
As a result, the directivity of the antenna generally forms a strong beam in the direction of the desired wave (wave from the mobile station (wireless communication apparatus) that desires communication) and forms a weak (or null) beam in the direction of the interference wave. .
In digital control, amplitude and phase can be controlled by multiplying complex numbers. This complex number is called a weight.

  As a weight calculation method, there are a ZF (Zero Forcing) method, an MMSE (Minimum Mean Square Error) method, and the like. In both methods, the weight is calculated using a known signal on the receiving side.

  The ZF method is a method of applying a weight obtained when noise is ignored and a composite output signal = a transmission signal. The ZF method shows good characteristics in an environment with few noises and interference waves, but the reception quality is significantly lowered as the interference wave component of the reception signal increases.

On the other hand, the MMSE scheme is a scheme that applies weights that minimize the error power of the combined output signal. Since the MMSE system has an interference wave removal capability, reception quality can be improved even when the interference wave component of the received signal is large.
As a calculation method of the MMSE method, there are LMS (Least Mean Square) based on the steepest descent method, RLS (Recursive Least Square), SMI (Sample Matrix Inversion) which is a direct solution method using sample values, and the like (see Non-Patent Document 2). ).
Note that Patent Document 1 discloses a technique for deriving a weight by the LMS algorithm in the OFDM system.

JP 2003-174427 A

IEEE 802.16 standard, “Part 16: AIR INTERFACE FOR BROADBAND WIRELESS ACCESS SYSTEMS”, 2007, USA Nobuyoshi Kikuma, "Adaptive signal processing by array antenna", first edition, Science and Technology Publishing Co., Ltd., 1998, p. 35-66

  In general, in order to derive an appropriate weight, it is necessary that a received signal includes a plurality of signals whose amplitude and phase at the time of transmission are known on the receiving side. On the receiving side, an appropriate weight can be derived from a plurality of received known signals by the MMSE method or the like.

Here, in the conventional weight derivation method, it is a natural premise that the plurality of known signals included in the received signal are transmitted from the same transmitter.
The appropriate weight varies depending on the transmission path environment between the receiver and the transmitter. Therefore, if a known signal transmitted from a completely different transmitter is included in the received signal, an appropriate weight cannot be derived. Because.

However, in WiMAX in which the above-described OFDMA scheme is adopted , since allocation is performed for a plurality of users, known signals (pilot subcarriers) transmitted from different transmitters are mixed in the received signal.
That, OFDMA scheme, like OFDM scheme which is a basic form, but has a plurality of sub-carriers in the frequency axis direction, the WiMAX OFDMA system is adopted, the frequency (subcarrier) and time ( A symbol is assigned to a plurality of users.
In WiMAX employing the OFDMA scheme, pilot subcarriers that are known signals are distributed and arranged in a two-dimensional subcarrier arrangement in the frequency axis direction and the time axis direction. These pilot subcarriers will be transmitted by different users.

As a result, in the WiMAX the OFDMA system is adopted, it becomes possible to perform weight derived using pilot subcarriers different user sends (known signal), it is difficult to correct weight deriving, painful suitable There was no specific knowledge on how to derive the weights.

Accordingly, the present invention is a communication system such as mobile WiMAX, and an object thereof is to derive the appropriate weight.

According to the present invention , wireless communication is performed by a communication method in which minimum unit areas for wireless resource allocation to users are arranged side by side in a frequency axis direction and a time axis direction so that the wireless resources can be shared by a plurality of users. A communication apparatus, comprising: a weight deriving unit for deriving a weight for a minimum unit area from reception signals of a plurality of subcarriers included in the minimum unit area .

According to the present invention, for deriving a weight in the minimum unit region of assignment of radio resources to the user, it can be derived applicable earnestly weights.
That is, if the weight is derived in the minimum unit of radio resource allocation to the user, it is guaranteed that the user is the same user within the minimum unit. Therefore, even if the user allocation information is not acquired, the user spans multiple users. Deriving weights can be prevented.

Another aspect viewed from inventions, a radio communication apparatus for performing wireless communication by OFDMA scheme, a plurality of received signal vectors as sample values, the correlation matrix and calculates a correlation vector, the calculated correlation matrix and correlation vector a SMI weight deriving unit that derives the SMI weight from, and a weight integration section for integrating the received signal vector a SMI weight derived by the SMI weight deriving unit, the SMI weight deriving unit includes a radio resource to the user This is a radio communication apparatus configured to derive an SMI weight for the minimum unit area using the received signal vectors of a plurality of subcarriers included in the minimum unit area of the allocation as the sample value. Thus, if the weight is derived in the minimum unit of radio resource allocation to the user, it is guaranteed that the same user is within the minimum unit. It is possible to prevent the weight from being derived.
Note that the derived weight may be a common weight in the region or a different weight for each subcarrier in the region.

Further, when the subcarrier arrangement of OFDMA is viewed in a two-dimensional arrangement in the frequency axis direction and the time axis direction, the minimum unit area is assigned to another minimum unit area adjacent in the frequency direction and the time axis direction, respectively. It is preferable that communication is performed by a communication method assigned to a user other than the user.
In such a case, since adjacent minimum units are assigned to different users, if a weight is derived in a range larger than the minimum unit, the weight is derived across multiple users. Since the weight is derived, this can be prevented.
The minimum unit area is preferably a tile in WiMAX uplink PUSC.

  A scalar matrix addition unit that adds a scalar matrix having a positive diagonal component to the correlation matrix calculated from the sample values, and derives an SMI weight using the correlation matrix to which the scalar matrix is added. Preferably, it is configured. When the SMI weight is derived using the correlation matrix, the inverse matrix operation of the correlation matrix is required. However, by adding a positive scalar matrix to the correlation matrix, the value of the correlation determinant becomes a certain value or more. Therefore, an overflow during calculation processing can be avoided, and the calculation is stabilized.

The sample value of the received signal vector preferably includes a received signal vector of a pilot subcarrier included in the user allocation area ( minimum unit area for allocation of radio resources to users ).
In addition, the SMI weight deriving unit calculates the user allocation area ( minimum unit area for allocating radio resources to users ) from the pilot subcarriers included in the user allocation area ( minimum unit area for allocating radio resources to users ). A transmission channel estimation unit that calculates a transmission channel estimation value of the data subcarrier included in the data subcarrier, and generates a reception signal vector of the data subcarrier from the transmission channel estimation value calculated by the transmission channel estimation unit, and data The SMI weight is preferably derived from the sample value including the received signal vector of the subcarrier. In this case, since the received signal vector of the data subcarrier can also be used, it is easy to increase the number of sample values.

The SMI weight derivation unit includes a received signal vector of a pilot subcarrier included in a user allocation area ( minimum unit area for radio resource allocation to a user ) and a user allocation area ( minimum unit for radio resource allocation to a user). Preferably, the SMI weight is derived using the received signal vector of the data subcarrier generated from the transmission path estimation value of the data subcarrier included in (region) as the sample value. By using the received signal vectors of both pilot and data subcarriers, the number of sample values increases.

The SMI weight deriving unit is preferably configured to derive an SMI weight for each data subcarrier included in the user allocation area (minimum unit area for radio resource allocation to a user) . In this case, an SMI weight for each data subcarrier is obtained.

The SMI weight deriving unit regards the user allocation area ( minimum unit area for radio resource allocation to a user ) as a set of a plurality of small areas that partially overlap each other, and includes pilot subcarriers included in the small area A correlation matrix calculation unit that calculates a correlation matrix for each of the small regions using the received signal vector of the data subcarrier or the reception signal vector of the data subcarrier as a sample value. Of these, it is preferable that the SMI weight is derived using the correlation matrix corresponding to one or a plurality of small regions to which the data subcarrier to which the SMI weight is to be derived belongs.
In this case, the small area used for weight derivation differs for each data subcarrier, and it is possible to derive an appropriate weight.
The SMI weight deriving unit regards the user allocation area ( minimum unit area for radio resource allocation to a user ) as a set of a plurality of small areas that partially overlap each other, and includes pilots included in the small area A correlation vector calculation unit that calculates a correlation vector for each small region using a reception signal vector of a subcarrier or a reception signal vector of a data subcarrier as a sample value, and a correlation for each small region calculated by the correlation vector calculation unit Of the vectors, the SMI weight is preferably derived using the correlation vector corresponding to one or a plurality of small regions to which the data subcarrier to which the SMI weight is to be derived belongs.
Also in this case, the small area used for the weight derivation differs for each data subcarrier, and it is possible to derive an appropriate weight.

According to the present invention, it can be derived applicable earnestly weights.

It is a figure which shows the communication system of WiMAX. It is a figure which shows the subcarrier two-dimensional arrangement | positioning of the uplink PUSC of WiMAX. It is a figure which shows a tile structure. It is a figure which shows the pilot signal arranged on a time axis at the same frequency. It is a block diagram of the receiving part which concerns on 1st Embodiment. It is a block diagram of the receiving part which concerns on 2nd Embodiment. It is a block diagram of the receiving part which concerns on 3rd Embodiment. It is a block diagram of the receiving part which concerns on 4th Embodiment. It is a figure which shows the small area | region in a tile. It is a figure which shows a correspondence table. It is a block diagram of the receiving part which concerns on 6th Embodiment. It is a figure which shows a simulation result. It is a figure which shows a simulation result. It is a figure which shows a simulation result.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the present embodiment, WiMAX is described as an example of the communication method, but the present invention is not limited to this.

  FIG. 1 shows a communication system in WiMAX. As shown in FIG. 1, a plurality of base stations BS1 and BS2 are installed, and one base station (base station radio communication device) BS1 and BS2 is provided for each cell. Each base station BS1 and BS2 communicates with mobile stations (user radio communication apparatuses) MS1 and MS2 in the cell. Each base station BS1 and BS2 can simultaneously perform communication with a plurality of mobile stations in the cell.

  In this WiMAX uplink, the base station BS1 receives a signal from the mobile station MS1, but a signal from the mobile station MS2 communicating with the base station BS2 of another cell becomes an interference wave. In order to remove this interference wave, the base station BS1 performs a process of deriving a weight (details will be described later).

  In WiMAX, an OFDMA system, which is a type of frequency multiplexing system, is employed. The OFDMA scheme is a scheme in which the concept of a logical subchannel configured by a subset of each subcarrier is introduced into the OFDM scheme to expand the flexibility of radio resource allocation to user data. Note that OFDM is a communication method for transmitting digital information by applying modulation such as QAM modulation to a large number of carriers (subcarriers) arranged so as to be orthogonal on the frequency axis.

There are three types of OFDMA subcarriers: a data subcarrier (Data Sub-Carrier), a pilot subcarrier (Pilot Sub-Carrier), and a null subcarrier (Null Sub-Carrier).
The data subcarrier (data signal) is a subcarrier for transmitting data and a control message. The pilot subcarrier is a known signal (pilot signal) on the reception side and the transmission side.

  The null subcarrier is a subcarrier in which nothing is actually transmitted, and is a guard subband on the low frequency side (guard subcarrier), a guard subband on the high frequency side (guard subcarrier), and a DC subcarrier. It is comprised by the carrier (center frequency subcarrier). In the following description, null subcarriers are not considered.

FIG. 2 shows a two-dimensional arrangement of data subcarriers and pilot subcarriers for WiMAX uplink PUSC (Partial Usage of SubChannels) employing OFDMA. In FIG. 2, the horizontal axis is a time (symbol) axis, and the vertical axis is a frequency (subchannel) axis.

In the uplink PUSC shown in FIG. 2, a total of 12 subcarriers, 3 in the symbol direction (time axis direction) × 4 in the frequency axis direction, are sub-set to form a tile T1, T2, T3 structure. . Each of the tiles T1, T2, and T3 is an area that is a minimum unit in user allocation.
Pilot subcarriers (black circles in FIG. 2) are arranged at the four corners of each tile T1, T2, T3, and other subcarriers in the tiles T1, T2, T3 are data subcarriers (white circles in FIG. 2). Has been.
As shown in FIG. 2, the tiles T1, T2, T3 are regularly arranged in the time axis direction and the frequency axis direction.

  In the following, for convenience, as shown in FIG. 3, four pilot subcarriers in one tile are indicated by “A, B, C, D”, respectively, and eight data subcarriers in the tile are respectively indicated. It shall be indicated by “1, 2, 3, 4, 5, 6, 7, 8”.

As described above, in WiMAX employing OFDMA, it is possible to allocate one communication frame (uplink subframe) to a plurality of users and share radio resources among a plurality of users.

WiMAX (uplink) user allocation employing OFDMA is performed in units of tiles ( minimum unit of allocation of radio resources to users ). An area in which a plurality of tiles ( minimum unit for assigning radio resources to users ) is combined is assigned to each user as a data area (burst area).

  As a method of combining a plurality of tiles constituting an area (burst area) allocated to one user, various forms can be assumed. Here, as shown in FIG. 2, each tile (minimum unit area) is It is assumed that a user assigned to another tile (minimum unit area) adjacent in the frequency direction and the time axis direction is assigned to another user.

  That is, in FIG. 2, tiles adjacent in the time axis direction are tiles of different users (for example, the tile T1-1 is a first user and the tile T2-1 is a second user). Further, tiles adjacent in the frequency direction are also different tiles (for example, the tile T-1 is the first user and the tile T2-2 is the second user). When such user allocation is performed, an appropriate weight cannot be obtained by applying the conventional weight derivation method as it is.

  That is, in order to derive weights in the OFDM system conventionally employed in digital terrestrial broadcasting (of course, user assignment is not performed), as shown in FIG. 4, a time axis at a certain frequency f (subcarrier) is used. A plurality of pilot subcarriers P1, P2, P3, P4,. Here, the reason why pilot subcarriers P1, P2, P3, P4,... Having the same frequency are used is that the optimum weights are considered to be different because the transmission path characteristics are different if the frequencies are different.

  In order to calculate a weight using, for example, the LMS algorithm using these pilot subcarriers P1, P2, P3, P4,..., First, the weight is updated using the pilot subcarrier P1, and then P2 , P3, P4,... (In order of time), the weights are sequentially updated (see Patent Document 1). By updating the weight many times, the weight converges to the optimum value.

  In addition to LMS, if SMI or RLS is simply adopted in the OFDMA scheme, a weight is obtained using a plurality of pilot subcarriers P1, P2, P3, and P4 arranged in the time axis direction at the same frequency as in LMS. Will be derived.

  Thus, in the conventional concept of deriving weights using a plurality of pilot subcarriers P1, P2, P3, and P4 arranged in the time axis direction at the same frequency, there is a possibility that the user may be switched in the time axis direction. A weight is derived using pilot subcarriers transmitted from different users, and an appropriate weight cannot be obtained.

In particular, as shown in FIG. 2, when the tiles adjacent to each other in the time axis direction are communication systems in which user allocation is performed such that tiles of different users are always tiles of different users, the base station BS1 Only a maximum of two can be acquired continuously in the time axis direction .

Therefore, in this embodiment, in order to derive the weight, not only the plurality of pilot subcarriers P1, P2, P3, and P4 arranged in the time axis direction at the same frequency are used, but the entire user allocation region or one of the user allocation regions is used. The weights for data subcarriers included in all or a part of the user allocation area are derived using (a part or all of) the pilot subcarriers included in the partial area.
Since the user who transmitted the pilot subcarrier is the same within one user allocation area, it is appropriate even if there is user allocation by acquiring the pilot subcarrier used for weight derivation from within the user allocation area. The weight can be derived.

In the following embodiments (first to fifth embodiments), as an example of using pilot subcarriers included in a part of the user allocation area, a minimum unit area for allocating radio resources to users A method using subcarriers included in a tile area will be described.

  When deriving weights in tile units as in the embodiment described below, instead of using only a plurality of pilot subcarriers arranged in the time axis direction at the same frequency, pilot subcarriers A and B in FIG. Pilot subcarriers at different positions in the frequency axis direction, such as C and D, are used for weight derivation.

In general, it is considered that the weight optimum value is different for different frequencies, and thus it is considered that received signals having different frequencies should not be used for weight derivation. However, if the frequency difference is as small as one tile, there is no problem.
Also, in the case of mobile communication such as WiMAX, when the moving speed of the mobile station MS is high, even if it is a subcarrier of the same frequency, the transmission path varies greatly with the passage of time, and the optimum weight value is also high. fluctuate. In this case, the use of a plurality of reception signals arranged in the frequency direction at the same time is smaller than that of a plurality of reception signals arranged in the time axis direction at the same frequency, and a more appropriate weight can be derived.

Therefore, in this embodiment, received signals (subcarriers) having different frequencies are positively used for weight derivation. In other words, in the conventional weight derivation, there was no idea of deriving weights using subcarriers included in a region that is spread in the frequency axis direction (regions for multiple subcarriers instead of one subcarrier). In the embodiment, weights are derived in units of areas that are spread in the frequency direction (and in the time axis direction), from the user allocation area (burst area) to the minimum unit area (tile area) for radio resource allocation to users .

  In particular, when deriving weights in units of tiles, even if they are at different positions in the frequency axis direction, such as pilot subcarriers A and B and C and D in FIG. There is not much influence due to the difference in frequency, and rather, since many pilot subcarriers can be used, an appropriate weight can be derived.

[First Embodiment]
FIG. 5 shows the configuration of the receiving unit 1 (first embodiment) of the base station communication apparatus that derives weights in tile units. In the case of OFDMA, the transmission side places data to be transmitted on frequency domain subcarriers, converts the frequency domain signal into a time domain signal by IDFT (Inverse Discrete Fourier Transform), and transmits the signal.

For this reason, the receiving unit 1 includes preprocessing units 12a and 12b that perform preprocessing such as frequency conversion and A / D conversion on the received signals received by the plurality of antenna elements 11a and 11b. The preprocessing units 12a and 12b convert the received signals into discrete baseband signals.
The received signal converted into the baseband signal is removed by CP (Cyclic Prefix) of the OFDM signal by the CP removal & DFT units 13a and 13b, and the time domain signal is converted into the frequency domain signal by DFT (Discrete Fourier Transform). Processing up to DFT is performed for each system of the antenna elements 11a and 11b.

The received signal in the frequency domain is given to a weight deriving unit 14 as an adaptive processor and a weight integrating & combining unit (weight integrating unit; weight combining unit) 15. The weight deriving unit 14 derives a weight (weight vector) for each antenna element in units of tiles using pilot subcarriers included in the received signal.
The weight integrating unit & combining unit 15 adds the weights to the data subcarriers included in the received signals at the antenna elements 11a and 11b, and combines the received signals obtained by integrating the weights. This integration and composition is also performed in tile units.
The combined received signal is demodulated by the demodulator 16 in tile units.

The weight deriving unit 14 is configured as an SMI weight deriving unit for deriving SMI weights. The SMI directly calculates the weight from the received signal and the known pilot signal (reference signal) instead of calculating the optimum weight value by the sequential update process as in the LMS and RLS. Since SMI is a kind of MMSE method, it has interference wave removal capability and can improve reception quality even if the interference signal component of the received signal is large.

In this embodiment, weights are derived in units of tiles. However, when sequential update processing is performed as in LMS or RLS, many weight updates are required before the weights converge to an optimum value .
Meanwhile, SMI are the method of calculating the weights directly from a received signal (and the reference signal), the sequential update process Ru unnecessary der.

Since the SMI weight is calculated from the correlation matrix and the correlation vector, the SMI weight derivation unit 14 includes a correlation matrix calculation unit 14a and a correlation vector calculation unit 14b.
The correlation matrix is calculated from a plurality of received signal vectors generated from the received signals received by the plurality of antenna elements 11a and 11b. The received signal vectors X _A , X _B , X _C , and X _D used for calculating the correlation matrix correspond to pilot subcarriers A, B, C, and D from within each tile in the received signals of the antenna elements 11a and 11b. Only the subcarrier to be extracted is generated.

That is, the received signal vector X _m (m = A, B, C, D) of each pilot subcarrier A, B, C, D is expressed as the following equation (1). Here, T represents transposition. N is the number of antenna elements.

The SMI weight deriving unit 14 includes in-tile pilot extracting units 14c and 14d for each of the antenna elements 11a and 11b in order to extract pilot subcarriers in each tile from the received signal.
For example, the in-tile pilot extraction unit of the N-th antenna element system includes four pilot subcarriers X _NA among the received signals X _Nm (m = A to D, 1 to 8) received by the N-th antenna element. , X _NB , X _NC , X _ND are extracted, and a set {X _NA , X _NB , X _NC , X _ND } consisting of these signals is generated.
The in-tile pilot extraction units of other antenna element systems also perform the same processing.

The correlation matrix calculation unit 14a receives pilot subcarrier sets {X _1A , X _1B , X _1C , X _1D } _,. X _NA , X _NB , X _NC , X _ND } (N is the number of antenna elements)), and the received signal vector X _m of the above equation (1) is obtained for each pilot subcarrier A, B, C, D in the tile. To generate.

Then, correlation matrix calculating unit 14a, from the reception signal vector X _m, based on the following formula (2), and generates a correlation matrix R for calculating the weight of the tiles. In the following formula (2), E [] represents expected value calculation, and H represents conjugate transposition.

Similarly to the correlation matrix calculation unit 14a, the correlation vector calculation unit 14b also receives the received signal vector X _m of the above formula (1) based on the pilot subcarriers acquired from the in-tile pilot extraction units 14c and 14d of each system. Is generated for each pilot subcarrier A, B, C, and D in the tile. In the first embodiment, these received signal vector X _m is a sample value for deriving the SMI weight for that tile.

Then, correlation vector calculating unit 14a, a reception signal vector X _m is a sample value, and a pilot signal known pilot signal generated by the generation unit 14e (the reference signal) S _m, in order to calculate the weight of the tile Is calculated based on the following equation (3). In the following formula (3), * represents a complex conjugate.

The SMI weight deriving unit 14 includes a weight calculating unit 14f that calculates an SMI weight (weight vector) for the tile using the correlation matrix and the correlation vector calculated based on the equations (2) and (3). Yes. The weight calculation unit 14f calculates the SMI weight W _SMI by the following equation (4).

The weight integrating & combining unit 15 uses the SMI weight vector W _SMI and the received signal vector X _m (m = 1 to 8) of the data subcarrier in the tile based on the following equation (5) to generate the combined output signal Y _m. Calculate Note that the data subcarriers to which weights are integrated are data subcarriers included in a tile having pilot subcarriers used for weight calculation.

In the first embodiment, since processing (from weight derivation to signal synthesis) is performed in units of tiles as described above, information on which user's tile is the target tile is unnecessary. In addition, processing is facilitated by applying a common SMI weight to each received signal vector in the tile.
In the above description, all four pilot subcarriers in the tile are used to derive weights. However, some subcarriers, for example, two or three subcarriers may be used.

[Second Embodiment]
FIG. 6 shows the receiving unit 1 of the base station communication apparatus according to the second embodiment. The second embodiment is different from the first embodiment in that a scalar matrix addition unit 14g is provided. In addition, the point which abbreviate | omitted description in 2nd Embodiment is the same as that of 1st Embodiment.

The scalar matrix adding unit 14g is a scalar matrix δI (I is a unit matrix) having a diagonal component having an arbitrary size with respect to the correlation matrix R calculated by the correlation matrix calculating unit 14a. , Δ is a positive real number), and a correlation matrix R ′ is calculated as in the following formula (6).

Then, the weight integration & synthesis unit 15 of the second embodiment calculates the SMI weight W _SMI based on the following equation (7) using the correlation matrix R ′ added with the scalar matrix δI.

  As described above, if a scalar matrix is added to the correlation matrix R, the value of the determinant necessary for the inverse matrix calculation of the correlation matrix performed when calculating the weight can be increased to some extent. It is possible to avoid a decrease in overflow and the like and stabilize the arithmetic processing. The magnitude of δ is preferably about several percent of the magnitude of the correlation matrix R.

[Third Embodiment]
FIG. 7 shows the receiving unit 1 of the base station communication apparatus according to the third embodiment. The third embodiment is different from the first embodiment, the correlation matrix calculating unit 14a and correlation vector calculating unit 14b is, as a received signal vector X _m, rather than using only pilot subcarriers (to D), Data subcarriers (1-8) are also used. In order to use data subcarriers, the SMI weight deriving unit 14 of the second embodiment includes transmission path estimation units 14h and 14i for each system of the antenna elements 11a and 11b. In addition, the point which abbreviate | omitted description in 2nd Embodiment is the same as that of 1st Embodiment.

The processing contents of the transmission path estimation units 14h and 14i are as follows. That is, each channel estimation unit 14h, 14i are tile pilot extracting unit 14c, and a reference signal S _m which is generated by the pilot subcarrier and the pilot signal generation unit 14e of the tile extracted by 14d, pilot subcarriers _Is calculated based on the following equation (8).

Furthermore, the transmission path estimation units 14h and 14i use the transmission path estimation values H _nm (n = 1,..., N) (m = A, B, C, D) of pilot subcarriers (A to D). Then, the data subcarrier transmission path estimation value H _nm (n = 1,..., N) (m = 1,..., 8) is obtained by an appropriate interpolation method such as linear interpolation.

と表すことができる。 The transmission path estimation value H _nm (n = 1,..., N) (m = 1,..., 8) of the data subcarrier obtained by interpolation transmitted “1” as the transmission signal S _m . Can be regarded as a received signal.
Therefore, in the case where “1” is transmitted as the transmission signal S _m in the data subcarriers (1 to 8), the reception signal of the data subcarrier is

It can be expressed as.

と表せる。
そこで、相関行列計算部１４ａ及び相関ベクトル計算部１４ｂは、伝送路推定部１４ｈ，１４ｉから取得したデータサブキャリアの伝送路推定値Ｈ_ｎｍ（ｎ＝１，・・・，Ｎ）（ｍ＝１，・・・，８）から、データサブキャリア（１〜８）の受信信号ベクトルを生成する。 Then, the received signal vector in the data subcarriers (1 to 8) is

It can be expressed.
Accordingly, the correlation matrix calculation unit 14a and the correlation vector calculation unit 14b transmit the transmission channel estimation values H _nm (n = 1,..., N) (m = 1) of the data subcarriers acquired from the transmission channel estimation units 14h and 14i. ,..., 8) generate reception signal vectors of data subcarriers (1 to 8).

  Then, the correlation matrix calculation unit 14a and the correlation vector calculation unit 14b calculate the correlation matrix and the correlation vector from the reception signal vectors of the pilot subcarriers (A to D) and the reception signal vectors of the data subcarriers (1 to 8). calculate.

なお、上記式（１１）（１２）において、パイロットサブキャリアにおける受信信号ベクトルについての下付添字ｍは、Ａ〜Ｄをとり、データサブキャリアにおける受信信号ベクトルについての下付添字ｍは、１〜８をとる。 That is, the formulas for obtaining the correlation matrix and correlation vector when using all 12 subcarriers combined with pilot subcarriers and data subcarriers in one tile are as shown in the following formulas (11) and (12). is there. The pilot signal (reference signal) corresponding to the data subcarriers (1 to 8) is “1”.

In the above equations (11) and (12), the subscript m for the received signal vector in the pilot subcarrier takes A to D, and the subscript m for the received signal vector in the data subcarrier is 1 to D. Take 8.

  According to the method of the third embodiment, since the number of received signal vectors as sample values increases, if the accuracy (interpolation accuracy) of the channel estimation value in the data subcarrier is high, the correlation matrix R and the correlation vector r Increases accuracy. As a result, the accuracy of the derived SMI weight is increased, and a more appropriate combined output signal can be obtained.

In the above, the transmission channel estimation values of all twelve data subcarriers in the tile are used, but it is not necessary to use all, and the transmission channel estimation values of some data subcarriers in the tile are used. May be.
In addition, the reception signal vector serving as the sample value may be a plurality of reception signal vectors of the pilot subcarrier and the data subcarrier, and any one of the reception signal vectors may be provided.

[Fourth Embodiment]
FIG. 8 shows the receiver 1 of the base station communication apparatus according to the fourth embodiment. In the fourth embodiment, the scalar matrix adder 14g of the second embodiment shown in FIG. 6 is added to the receiver 1 of the third embodiment shown in FIG. Since other points are the same as those of each embodiment, description thereof is omitted here.

[Fifth Embodiment]
In the fifth embodiment, the basic configuration of the receiving unit 1 is the same as that of the third embodiment shown in FIG. 7, and therefore the fifth embodiment will be described with reference to FIG. However, in the fifth embodiment, instead of deriving a common weight _WSMI in one tile as in the first to fourth embodiments, each individual data subcarrier (1-8) in the tile In addition, the weights W1 _{SMI to} W8 _SMI are derived individually.

That is, the correlation matrix calculation unit 14a and the correlation vector calculation unit 14b of the fifth embodiment do not calculate one correlation matrix and correlation vector for the entire tile, but as shown by the six squares in FIG. Correlation matrices R _{1 to} R ₆ and correlation vectors r _{1 to} r ₆ are calculated for each of a plurality of (six) small areas within a tile ( minimum unit for assigning radio resources to users ).

For example, the correlation matrix calculation unit 14a and the correlation vector calculation unit 14b relate to the first small region, the received signal vector X _A of the pilot subcarrier _A and the data subcarriers 1, 3, and 4 in the first region. Correlation matrix R ₁ and correlation vector r ₁ are calculated using received signal vectors X ₁ , X ₃ , and X ₄ . Similarly, correlation matrices and correlation vectors are calculated for other small regions.
As shown in FIG. 9, the small areas in the tile are set so as to partially overlap each other, but the size and number are not particularly limited. Further, the size may be different for each small region.

Weight calculation unit 14f, using the correlation matrix _R 1 to R ₆ and the correlation vector _r 1 ~r ₆ of each small region, based on the correspondence table shown in FIG. 10, SMI weight for each data subcarrier (1-8) Calculate W _SMI .
The correspondence table in FIG. 10 shows the sum of correlation matrices and the sum of correlation vectors of one or more small regions to which each data subcarrier in the tile belongs.
For example, in the table of FIG. 10, the correlation matrix of the data subcarrier “1” is “the sum of the correlation matrix R ₁ of the first small region and the correlation matrix R ₂ of the second small region (R ₁ + R ₂ )”. It has become. This is because, in FIG. 9, the data subcarrier “1” belongs to the first small area and the second small area. Similarly, the correlation vector of data subcarrier “1” is (r ₁ + r ₂ ).

Therefore, the SMI weight W1 _SMI for the data subcarrier “1” is obtained by substituting the correlation matrix (R ₁ + R ₂ ) into R in the equation (4) and replacing the correlation vector (r ₁ + r ₂ ) in the equation (4). It is calculated by substituting for r.
Similarly, SMI weights W2 _{SMI to} W8 _{SMI of} other data subcarriers (2 to 8) are also calculated.
Then, the weight integrating & synthesizing unit 15 integrates the weights W1 _{SMI to} W8 _SMI respectively corresponding to the individual data subcarriers (1 to 8) to obtain a combined output signal.

  According to the fifth embodiment, the reception vector serving as the sample value differs for each data subcarrier, and the reception signal vector near each data subcarrier is used for weight derivation. If the estimation accuracy is high, the weight can be derived with high accuracy.

[Sixth Embodiment]
FIG. 11 shows the receiving unit 1 of the base station communication apparatus according to the sixth embodiment. In the sixth embodiment, the entire individual user allocation area is used as a unit for deriving SMI weights, not the minimum unit of radio resource allocation to users called tiles (part of the user allocation area).
Therefore, UL-MAP (UpLink-MAP) information is given to the correlation matrix calculation unit 14a, the correlation vector calculation unit 14b, and the weight integration & synthesis unit 15 as user allocation information. This UL-MAP information is generated by the base station communication apparatus, and is used to notify the allocated area to all mobile stations on the downlink (downlink frame). The receiving unit 1 of the base station uses the UL-MAP information to grasp which area (burst area) in the uplink frame is assigned to which user, and generates an SMI weight for each burst area. Performs weight accumulation / synthesis.

  Note that the UL-MAP information is used not only when the entire user allocation area is a unit for deriving SMI weights, but also by using a partial area of the user allocation area as a unit for deriving SMI weights. Even when the area is larger than the user-assigned minimum unit area, it can be used to determine the range of the same user.

  11 shows the same basic configuration as that of the receiving unit 1 of the first embodiment shown in FIG. 5, but the concept of the sixth embodiment can also be applied to the configurations of the second to fourth embodiments. It is.

[simulation result]
FIGS. 12 and 13 show a third embodiment (hereinafter referred to as “SMI-1”) in which all subcarriers (12) in a tile are sample values of a received signal vector, and four pilot subcarriers in a tile. Each of the first embodiment (hereinafter referred to as “SMI-2”) that is a sample value of a received signal vector and the fifth embodiment (hereinafter referred to as “SMI-3”) that derives a weight for each data subcarrier. The result of comparing the performance by simulation is shown.

In the simulation results shown in FIG. 12 and FIG. 13, the combined output signal for each CNR (Carrier-to-Noise Ratio) and CIR (Carrier-to-Interference Ratio) in each of the SMI-1, SMI-2, and SMI-3 systems. Compared with the CINR (Carrier-to-Interference-plus-Noise Ratio) distribution of
The parameters for this simulation are as follows.
CNR: 0 [dB], 10 [dB], 20 [dB], 30 [dB]
CIR: 0 [dB], 20 [dB]
Slot number: 10,000 (= 60,000tiles)
Data subcarrier modulation method: QPSK

  Note that the mobile stations are a mixture of low speed (for example, 30 km / h) and high speed (for example, 120 km / h).

12 and 13 show a cumulative distribution function of CINR. The horizontal axis is CINR, and the vertical axis is the probability that the combined output signal is lower than the CINR.
FIG. 12 shows a case where CIR = 0 [dB], that is, a case where the desired wave and the interference wave have the same magnitude. In this case, it can be seen that when the CNR is bad (CNR = 0 [dB]), a difference appears in each method.
SMI-2 shown in FIG. 12B is the best characteristic, and then the characteristics deteriorate in the order of SMI-1 shown in FIG. 12A and SMI-3 shown in FIG.

  When the interference wave is large, the interpolation accuracy of the transmission path estimation value of the data subcarrier decreases. In such a case, SMI-2 using only the received signal vector of the pilot subcarrier as a sample value is good.

  FIG. 13 shows a case where CIR = 20 [dB] and the desired wave is larger than the interference wave, and the accuracy of the transmission path estimation for the data subcarrier is low. In this case, as in FIG. 12, SMI-2 shown in FIG. 13B is the best characteristic, followed by SMI-1 shown in FIG. 13A and SMI-3 shown in FIG. In this order, the characteristics deteriorate.

FIG. 14 shows a case where CIR = 20 [dB] and the accuracy of channel estimation for data subcarriers is high. In this case, SMI-3 shown in FIG. 14 (c) has the best characteristics, and then the characteristics are worse in the order of SMI-1 shown in FIG. 14 (a) and SMI-2 shown in FIG. 14 (b). Become.
As shown in FIG. 14, if the accuracy of channel estimation is high, the characteristics of SMI-1 and SMI-3 using the received signal vector of the data subcarrier are improved, and in particular, the SMI for deriving the weight for each data subcarrier. -3 characteristics are excellent.

The present invention is not limited to the embodiment described above, and various modifications can be made without departing from the spirit of the present invention. For example, the minimum unit of radio resource allocation to a user is not limited to the tile shown in FIG. 3, and may be a downlink PUSC cluster, for example.

  1: receiving unit 11a, 11b: antenna element 12a, 12b: preprocessing unit 13a, 13b: CP removal & DFT unit 14: SMI weight deriving unit 15: weight integrating & combining unit (weight integrating unit) 16: demodulating unit

Claims (1)

A wireless communication apparatus that performs wireless communication using an OFDMA scheme,
A SMI weight deriving unit for calculating a correlation matrix and a correlation vector using a plurality of received signal vectors as sample values, and deriving an SMI weight from the calculated correlation matrix and the correlation vector;
A weight integrating unit that integrates the SMI weight derived by the SMI weight deriving unit into a received signal vector;
With
The SMI weight deriving unit is configured to derive an SMI weight for the user allocation region using the received signal vectors of a plurality of subcarriers included in the user allocation region in the OFDMA scheme as the sample values. A wireless communication device.

Images