JP4476879B2 - Spatial multiplexing transmission receiving method and apparatus - Google Patents

Description

  The present invention relates to a spatial multiplexing transmission receiving method and apparatus for receiving a plurality of different signals at the same time using a plurality of antenna elements.

A conventional receiver using the orthogonal wave frequency division multiplexing system is a receiver that decodes a transmission signal multiplexed in the frequency domain by receiving a signal with one antenna element.
FIG. 5 shows the configuration of a receiving apparatus using a conventional orthogonal frequency division multiplexing system. Reference numeral 900 denotes an antenna element, 910 denotes an AGC (Automatic Gain Control), 920 denotes an A / D (analog / digital) converter, 930 denotes a signal acquisition point determination device, 940 denotes an FFT (Fourier transform) device, and 950 denotes a transfer coefficient estimation. 960 is a transfer coefficient equalizing device, 970 is a known signal phase detecting device, 980 is a phase rotation amount correcting device, and 990 is a subcarrier decoding device.

  The gain received by the AGC 910 is set by the AGC 910 so that the signal received by the antenna element 900 falls within the dynamic range, is converted to a digital signal by the A / D converter 920, and the signal acquisition point determination device 930 Is output. The signal acquisition point determination device 930 obtains the start point of the orthogonal frequency division multiplexing (OFDM) symbol from the known signal for sampling synchronization, removes the guard interval, and outputs it to the FFT device 940. The FFT device 940 converts a time-series signal into a frequency component signal by Fourier transform and outputs the signal to the transfer coefficient estimation device 950. The transfer coefficient estimation device 950 estimates the transfer coefficient from the input signal for the transfer coefficient estimation known signal, and outputs the transfer coefficient of each frequency band and the received signal to the transfer coefficient equalization device 960. The transfer coefficient equalizer 960 corrects the transmission path distortion using the transfer coefficient of the frequency band corresponding to the received signal, and the corrected received signal is output to the phase rotation amount correcting device 980, and is transmitted to a specific frequency band. The included known signal is output to the known signal phase detector 970. The known signal phase detection device 970 calculates a phase rotation amount and outputs it to the phase rotation amount correction device 980 in order to follow a phase that rotates with time using a signal that is known in the reception device. The phase rotation amount correction device 980 corrects the received signal input from the transfer coefficient equalization device 960 using the phase rotation amount input from the known signal phase detection device 970 and outputs the correction signal to the subcarrier decoding device 990. I do. Subcarrier decoding apparatus 990 can decode a signal obtained in each frequency band and obtain a signal transmitted in the transmission apparatus.

Hereinafter, the contents of signal processing in the above configuration will be described in detail. Here, consider that performs communication by F-number of frequency division multiplexing, the i-th transmission coefficient representing the propagation environment of the band (1 ≦ i ≦ F) between the transmitter and the receiver and H _i . The transfer coefficient estimation device 950 estimates the transfer coefficient from the transfer coefficient estimation known signal included in the received signal. When the estimated transfer coefficient is expressed as H ₁ ′ to H _F ′, the phase of H _i ′ is rotated by the phase rotation amount θ (t ₀ ) at the transfer coefficient acquisition timing t ₀ , and H _i ′ is H _i ′. Using

と表すことができる。
一方、受信装置において受信され、フーリエ変換（ＦＦＴ）されたｉ番目のサブキャリアの受信タイミングｔにおける受信信号Ｘ_ｉは、送信信号Ｓ_ｉおよび伝達係数Ｈ_ｉを用いて、

It can be expressed as.
On the other hand, the reception signal X _{i at} the reception timing t of the i-th subcarrier received by the reception device and subjected to Fourier transform (FFT) is transmitted using the transmission signal S _i and the transfer coefficient H _i .

と表すことができる。すると、伝達係数等化装置９６０では、伝達係数推定装置９５０で推定されたＨ_ｉ’（（１）式）および受信信号Ｘ_ｉ（（２）式）より推定送信信号Ｓ_ｉ’を以下のように得ることができる。

It can be expressed as. Then, the transmission coefficient equalizer 960 obtains the estimated transmission signal S _i ′ from H _i ′ (Equation (1)) and the received signal X _i (Equation (2)) estimated by the transmission coefficient estimator 950 as follows. Can get to.

この推定送信信号Ｓ_ｉ’は（ｔ−ｔ_０）だけ位相が回転してしまい、伝送品質が著しく劣化している。これに対し、既知信号位相検出装置９７０は受信信号Ｘ_ｉのうち、特定の周波数帯（ｊ０）に既知信号Ｓ_ｊ０が含まれることを利用し、推定送信信号Ｓ_j0’を既知信号Ｓ_j0で除算することで、位相変化量θ（ｔ−ｔ_０）を

The estimated transmission signal S _i ′ rotates in phase by (t−t ₀ ), and the transmission quality is significantly degraded. In contrast, the known signal phase detector 970 in the received signal X _i, using the fact that contains known signals S _j0 to a particular frequency band (j0), the estimated transmission signal S _j0 'in the known signal S _j0 By dividing the phase change amount θ (t−t ₀ ),

として推定することができる。そこで、位相回転量補正装置９８０において、受信タイミングｔの推定受信信号Ｓ_１〜Ｓ_Ｆに対し、ｅｘｐ（−ｊθ（ｔ−ｔ_０））を乗算することで、位相回転を補償し、送信信号Ｓ_１〜Ｓ_Ｆを得ることができる。
なお、従来技術として非特許文献１が知られている。
守倉正博、久保田周治、「改訂版８０２．１１高速無線ＬＡＮ教科書」、２００４年、ｐ．１６５−１６７，２１５−２１９

Can be estimated as Therefore, the phase rotation amount correcting apparatus 980, to estimate the received signal _S 1 to S _F of the received timing t, by multiplying the _{exp (-jθ (t-t 0} )), to compensate for the phase rotation, the transmission signal it can be obtained S ₁ to S _F.
Note that Non-Patent Document 1 is known as a prior art.
Masahiro Morikura, Shuji Kubota, “Revised 802.11 High-Speed Wireless LAN Textbook”, 2004, p. 165-167, 215-219

Although the above technique enables communication in orthogonal frequency division multiplexing, a signal sequence received at the same time includes a synchronization error such as phase noise, or when receiving a signal from a different transmitter. Interference between signals becomes a problem, and a shift in estimation timing of transfer coefficients and a phase shift due to insufficient synchronization between transmitters.
The present invention has been made in view of such circumstances, and uses space division multiplexing to compensate for phase rotation even when there is inter-signal interference and a phase shift occurs due to a shift in the estimation timing of the transfer coefficient. It is an object of the present invention to provide a spatial multiplexing transmission receiving method and a spatial multiplexing transmission receiving apparatus that enable this.

  The present invention has been made to solve the above problems, and the invention according to claim 1 is provided with a plurality of antenna elements in a receiving apparatus, using an orthogonal wave frequency division multiplexing system, at the same time and at the same frequency. A method for compensating for phase rotation of signal points at the time of receiving and decoding signal sequences transmitted from a plurality of transmitters, receiving known signals transmitted from a transmitter at a plurality of transmission timings, and transmitting devices Calculating a first-order estimated transfer coefficient matrix that estimates a transfer coefficient matrix including a phase shift between the receiver and the reception device, decoding the received signal using the first-order estimated transfer coefficient matrix, and including in the decoded signal Calculating the instantaneous phase correction amount of all signal sequences from known signals used in a specific frequency band of the orthogonal frequency division multiplexing received at the same time or a plurality of times, and using the instantaneous phase correction amount It is a spatial multiplexing transmission reception method characterized by comprising the step of correcting the time variation of the phase rotation of the signal sequence.

  The invention according to claim 2 is included in the reception signals of a plurality of reception timings in the spatial multiplexing transmission reception method according to claim 1 when the simultaneously received signal sequences are not completely synchronized. A phase rotation amount of each signal series is calculated from a known signal, a difference from the instantaneous phase correction amount of the phase rotation amount of each signal series is calculated as a phase rotation correction amount, and each signal is calculated using the phase rotation correction amount. It has the process of performing the phase rotation correction of the series.

  According to the third aspect of the present invention, transmission signals of N signal sequences transmitted at the same time using an orthogonal wave frequency division multiplexing system are transmitted by M received signals obtained from M antenna elements at the same time. In a receiving apparatus for spatial multiplexing transmission that performs decoding, an A / D conversion apparatus that performs analog / digital conversion connected to each antenna element, and a sampling synchronization of the obtained data that is connected to the A / D conversion apparatus A signal acquisition point determination device that determines an acquisition point of an orthogonal frequency division multiplexing OFDM signal from a known signal, an FFT device that performs Fourier transform on the OFDM signal and calculates a signal in each frequency band, and the FFT A transmission coefficient matrix primary estimation unit that is connected to a device and calculates a primary estimation transmission coefficient matrix from an OFDM known signal for estimating a transmission coefficient matrix included in the OFDM signal Connected to the transmission coefficient matrix primary estimation device, and using the primary estimation transmission coefficient matrix, decodes M OFDM signals received at the same time, and decodes known signals included in each signal sequence A spatial division multiplex decoding apparatus that outputs a result and a decoding result of a desired data signal; and a decoding of a known signal of each signal series provided corresponding to each signal sequence and input from the spatial division multiplex decoding apparatus From the result, a phase rotation amount at a predetermined time of each signal sequence is calculated, and a known signal phase detection device that outputs the result, and a predetermined time of each input signal sequence connected to the known signal phase detection device The phase rotation amount of all signal sequences is calculated and output from the phase rotation amount at, and the phase rotation amounts at multiple symbol timings of each input signal sequence are averaged. A phase rotation correction amount calculating device for calculating a phase rotation amount of each signal sequence and outputting a difference between the calculated phase rotation amount of each signal sequence from the phase rotation amount of all the signal sequences; and the space division multiplexing The decoding result of the desired data signal input from the decoding device is based on the phase rotation amount of the entire signal sequence at the time corresponding to the decoding result of the desired data signal input from the phase rotation correction amount calculation device. A common phase rotation amount correction device that performs signal point correction and outputs a primary correction decoding result, and a primary correction decoding result input from the common phase rotation amount correction device is input from the phase rotation correction amount calculation device. Further, signal phase correction is performed based on the difference of the phase rotation amount of each signal sequence from the phase rotation amount of all the signal sequences, and a secondary correction decoding result is output; and Signal series And a subcarrier decoding device for calculating desired bit data from a signal point of an input secondary correction decoding result, connected to the interphase difference correction device, and a spatial multiplexing transmission receiving device .

  According to a fourth aspect of the present invention, in the spatial multiplexing transmission receiver according to the third aspect, the known signal phase detection device is configured to output all signal sequences determined from the phase rotation amounts of a plurality of input symbol timings. From the phase rotation amount, an estimated value of the phase change amount of time is calculated, and the phase rotation amount of all signal sequences at a predetermined symbol timing is determined so as not to cause a change in phase rotation amount over time, and the common phase rotation is performed. It outputs to a quantity correction apparatus, It is characterized by the above-mentioned.

  According to a fifth aspect of the present invention, in the spatial multiplexing transmission receiver according to the third aspect, the known signal phase detection device is configured to receive all signal sequences determined from the phase rotation amounts of a plurality of input symbol timings. Fitting a given function from the amount of phase rotation, calculating the amount of phase change over time, determining the amount of phase rotation of all signal sequences at a predetermined symbol timing so as not to cause phase rotation, and the common phase rotation It outputs to a quantity correction apparatus, It is characterized by the above-mentioned.

  According to a sixth aspect of the present invention, in the spatial multiplexing transmission receiving apparatus according to the third or fourth aspect, the spatial division multiplexing decoding apparatus uses the primary estimated transmission coefficient matrix to receive M received at the same time. Decode the OFDM signals, and output the decoding result of the known signal included in each signal sequence and the probability of the known signal given as an estimated value to the primary estimated transfer coefficient matrix to the known signal phase detection device In addition, the decoding result of the desired data signal is output to the common phase rotation amount correction device, and the known signal phase detection device weights or selects the decoding result of the known signal based on the decoding result of the input known signal and its probability. The phase rotation amount of each signal series is determined, the result and the probability are output to the phase rotation correction amount calculation device, and the phase rotation correction amount calculation is performed. The apparatus performs weighting or selection based on the phase rotation amount of each signal sequence input from the known signal phase detection device and its certainty, calculates the phase rotation amount of all signal sequences, and calculates the common phase rotation. The amount of phase rotation at each of a plurality of symbol timings of each signal sequence output to the amount correction device is weighted or selected based on the input probability, and the difference from the phase rotation amount of all the signal sequences Is calculated and output to the inter-signal sequence phase difference correction apparatus.

  A seventh aspect of the present invention is the spatial multiplexing transmission receiver according to any one of the third to fifth aspects, wherein the transmission represents an actual propagation environment including a case where a transmission weight that is optimum for the propagation environment is obtained. When a coefficient matrix is used, the primary estimated transfer coefficient matrix is corrected using the difference between the phase rotation amount of each signal sequence and the phase rotation amount of all signal sequences obtained by the phase rotation correction amount calculation device. The transfer coefficient matrix obtained by this is used.

  The invention according to claim 8 is the spatial multiplex transmission receiver according to any one of claims 3 to 5, wherein the transmission represents an actual propagation environment including a case where a transmission weight optimum for the propagation environment is obtained. When a coefficient matrix is used, a temporal phase change amount obtained from the phase rotation amount of all signal sequences obtained by the phase rotation correction amount calculation device and the primary estimated transfer coefficient matrix with respect to the primary estimated transfer coefficient matrix. A transfer coefficient matrix obtained by performing correction using the phase change amount obtained from the calculated symbol timing is used.

  According to the present invention, in a receiving apparatus, it is possible to accurately follow phase rotation with respect to signals of a signal sequence transmitted from a plurality of transmitting apparatuses at the same time, thereby preventing deterioration in transmission quality.

  Hereinafter, an embodiment of the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing a configuration of a spatial multiplexing transmission receiver according to an embodiment of the present invention. This receives signal sequences transmitted from a plurality of transmitters at the same time, estimates a transfer coefficient matrix including an initial phase shift from a plurality of known signals having different reception timings, and uses a plurality of received signals including inter-signal interference. 2 shows a configuration in which phase rotation is followed and transmission quality is compensated. Reference numerals 100-1 and 100-2 to 100-M denote antenna elements, 110-1 and 110-2 to 110-M denote AGCs, 120-1 and 120-2 to 120-M denote A / D converters, and 130 denotes Signal acquisition point determination device, 140 is an FFT device, 150 is a transfer coefficient matrix primary estimation device, 160 is a spatial division multiplexing decoding device, 170-1 to 170-N are known signal phase detection devices, and 171 is a phase rotation correction amount calculation device. , 180 is a common phase rotation amount correction device, 181-1 to 181-N are signal sequence phase difference correction devices, and 190 is a subcarrier decoding device. That is, M antenna elements, AGCs, and A / D converters are provided, and N known signal phase detectors and inter-sequence difference correction units are provided corresponding to the number of signal sequences to be transmitted. It has been.

  The gain received by the AGC 110-1 is set in the A / D converter 120-1 so that the signal received by the antenna element 100-1 enters the dynamic range, and is converted into a digital signal by the A / D converter 120-1. And output to the corresponding port of the signal acquisition point determination device 130. Similarly, for the signals received by the antenna elements 100-2 to 100-M, gain settings are performed in the AGs 110-2 to 110-M, respectively, and digital signals are respectively obtained by the A / D conversion devices 120-2 to 120-M. And output to the corresponding port of the signal acquisition point determination device 130.

  The signal acquisition point determination device 130 has M input ports and M output ports, obtains an OFDM symbol start point from a known signal for sampling synchronization, removes the guard interval, and removes the FFT device 140. Output to the corresponding port. The FFT device 140 converts a time-series signal received by each antenna element by Fourier transform into a frequency component signal, and outputs it to the transfer coefficient matrix primary estimation device 150. The transfer coefficient matrix primary estimation device 150 obtains a primary estimated transfer coefficient matrix including an initial phase rotation from known signals for transfer coefficient matrix estimation obtained at a plurality of reception timings among input signals, and this primary estimated transfer coefficient A received signal including a matrix and desired data is output to space division multiplex decoding apparatus 160. The spatial division multiplexing decoding apparatus 160 uses the input primary estimated transfer coefficient matrix and receives the same reception timing by a decoding method such as ZF (Zero Forcing), MMSE (Minimum Mean Square Error), or SUC (Successive Cancellation). The phase rotation decoded signal is calculated from the signal, the phase rotation decoded signal of the known signal included in the specific frequency band is output to the known signal phase detection devices 170-1 to 170-N, and the rest is corrected for the common phase rotation amount Output to device 180.

  The known signal phase detection devices 170-1 to 170-N calculate the phase rotation amount in each signal series from the known signal held by the receiving device and the phase rotation decoding result of the known signal, and the phase rotation correction amount calculation device It outputs to 171. The phase rotation correction amount calculation device 171 calculates the instantaneous phase correction amount for the entire signal sequence and the deviation from the instantaneous phase correction amount for each signal sequence. Output to the phase difference correction devices 181-1 to 181-N.

  The common phase rotation amount correction apparatus 180 uses the instantaneous phase correction amount in the entire signal sequence input from the phase rotation correction amount calculation apparatus 171 and uses the temporal variation of the phase rotation of the decoded signal input from the spatial division multiplexing decoding apparatus 160. And is output to the signal sequence phase difference correction devices 181-1 to 181-N as primary correction decoded signals. In the inter-signal-sequence phase difference correction devices 181-1 to 181-N, each signal sequence input from the phase rotation correction amount calculation device 171 is used as a phase rotation correction amount by using a deviation from the instantaneous phase correction amount in each signal sequence. Phase rotation correction is performed on the primary correction decoding signal of the series, and the secondary correction decoding signal is output to subcarrier decoding apparatus 190. In subcarrier decoding apparatus 190, the signal transmitted in the transmission apparatus can be obtained from the secondary correction decoded signal obtained in each frequency band and each signal series.

Hereinafter, the contents of signal processing in the above configuration will be described in detail. Considering communication using F frequency division multiplexes for transmission signals of N signal sequences transmitted at the same time, the i-th (1 ≦ i ≦ F) between the transmission device and the reception device. Hereinafter, the transmission coefficient matrix representing the propagation environment of each transmission signal sequence in the frequency band is denoted as H _i and is expressed as follows.

このとき、Ｈ_ｉ，ｊｋはｉ番目の周波数帯におけるｋ番目の信号系列のｊ番目の受信アンテナ素子における伝達係数を表す。

At this time, H _{i, jk} represents a transfer coefficient in the j-th receiving antenna element of the k-th signal sequence in the i-th frequency band.

In order to estimate this transfer coefficient matrix, it is necessary to transmit the known signal S ₀ for estimating the transfer coefficient matrix in all the transmission apparatuses. S ₀ is represented by a matrix of (number of transmission signal sequences) × (number of transmission signal sequences) or a matrix having more arrays. Hereinafter, for the sake of simplicity, S ₀ is assumed to be an N × N diagonal matrix. This corresponds to transmitting known signals at timings independent of each other in each transmitting apparatus.

The received signal received by each antenna element is selected from the optimum signal acquisition point via AGC 110-1, 110-2 to 110-M, A / D converter 120-1, 120-2 to 120-M. A received signal x _i (t) at a reception timing t in the i-th frequency band is obtained by performing Fourier transform on the time-series signal (M × 1 vector). The transfer coefficient matrix primary estimation device 150 obtains a primary estimated transfer coefficient matrix H _i ′ from received signals x _i (T ₁ ) to x _i (T _N ) of known signals for transfer coefficient estimation included in the received signal. Here, if the reception signal for propagation environment estimation is X _i ′ = (x _i (T ₁ ), x _i (T ₂ ),..., X _i (T _N )), the transfer coefficient matrix H _i is known. It can be expressed as follows using the signal S ₀ .

上記のＮ_ｉはＭ×Ｎの行列であり、熱雑音から構成される。Φはチャネル行列（伝達係数行列）の推定が複数の受信タイミングにおける信号から行われるために生じる初期位相回転により表せるＮ×Ｎの対角行列であり、

The above N _i is an M × N matrix and is composed of thermal noise. Φ is an N × N diagonal matrix that can be expressed by the initial phase rotation that occurs because the estimation of the channel matrix (transfer coefficient matrix) is performed from signals at a plurality of reception timings,

と表すことができる。Ｎ_ｉ’は上記の熱雑音から構成される行列Ｎ_ｉにΦが乗算されたものであるが、Φは位相を回転させる演算子なので、本質的にＮ_ｉと変わらない。
ここで受信信号Ｘ_ｉ’から既知信号を用いて伝達係数行列の推定を行うと、（６）式を用いて一次推定伝達係数行列は

It can be expressed as. N i _'are those [Phi is multiplied by a matrix N _i consists above thermal noise but, [Phi is because operator for rotating the phase unchanged essentially N _i.
Here, when the transfer coefficient matrix is estimated using the known signal from the received signal X _i ′, the first-order estimated transfer coefficient matrix is calculated using Equation (6).

と求めることができる。よって得られたチャネル行列はΦが乗算されていることにより、列成分に位相回転が生じている。

It can be asked. Thus, the obtained channel matrix is multiplied by Φ so that phase rotation occurs in the column component.

This phase rotation term uses, for example, the repetition portion of a known signal for determining the OFDM acquisition point, the fact that the actual transmission signal is repeated as a guard interval of the OFDM signal, etc., and rough estimation of the phase rotation amount, It can be reduced, but it cannot be completely removed. In the following, the thermal noise N _i ″ is assumed to be negligibly small for simplicity, and the decoding method of the desired data signal is shown for the case where the primary estimated transfer coefficient matrix H _i ′ is used and ZF is used as the decoding method.

A signal (N × 1 vector) transmitted by the transmission apparatus corresponding to the reception signal x _i (t) at the reception timing t in the i-th frequency band is defined as s _i (t). In the receiving apparatus, communication with high transmission quality can be realized by accurately restoring s _i (t). The received signal x _i (t) is obtained by using the transmission signal s _i (t) and the transfer coefficient matrix H _i.

と表すことができる。ここで、ｎ_ｉ(t)はＭ×１の熱雑音ベクトルである。一次推定伝達係数行列Ｈ_ｉ’の逆行列もしくは擬似逆行列（Ｎ≠Ｍの場合）をＨ_ｉ’^−１と表現することとすると、受信信号に一次推定伝達係数行列Ｈ_ｉ’から前述のＺＦによる復号を行うと、得られる復号信号ｓ_ｉ’(t)は、（８）式（ただし、Ｎ_ｉ’’は無視できるほど小さいものとする）、（９）式を用いて、

It can be expressed as. Here, n _i (t) is an M × 1 thermal noise vector. The primary estimated transfer coefficient matrix H _i if _'of the inverse matrix or the pseudo inverse matrix (N ≠ case of M) H _i' and be expressed as ^-1, described above from the primary estimated transfer coefficient matrix H _{i 'in} the received signal ZF When the decoding by s is performed, the obtained decoded signal s _i ′ (t) is obtained by using equation (8) (where N _i ″ is negligibly small) and equation (9):

と表すことができる。上式でｅｘｐ（ｊθ(t)）Φ^−１をΨ(t)とおいた。上式から分かるように、復号信号ｓ_ｉ’ (t)は位相回転の時変動ｅｘｐ（ｊθ(t)）と、初期位相回転による演算行列Φ^−１により表せる位相回転演算行列Ψ(t)（Ｎ×Ｎの対角行列）により信号位置が回転してしまっている。ここで、Ψ(t)の対角要素を（ψ_１(t)，ψ_２(t)，・・・，ψ_Ｎ(t)）とすると、ｋ番目の信号系列の位相回転量を示す演算子ψ_ｋは

It can be expressed as. In the above equation, exp (jθ (t)) Φ ⁻¹ is set as Ψ (t). As can be seen from the above equation, the decoded signal s _i '(t) is time variation of phase rotation exp (jθ (t)) and the initial phase rotation due expressed by calculating the matrix [Phi ^-1 phase rotation operation matrix Ψ (t) ( The signal position is rotated by (N × N diagonal matrix). Here, if the diagonal elements of ψ (t) are (ψ ₁ (t), ψ ₂ (t),..., Ψ _N (t)), the calculation indicating the phase rotation amount of the k-th signal sequence. The child ψ _k is

と表すことができる。

It can be expressed as.

Therefore, in order to correct this phase rotation, the known signal phase detection devices 170-1 to 170-N perform phase rotation correction using the known signal transmitted in a specific frequency band and its decoded signal. Here, the frequency band in which the known signal is transmitted is u-th, and the transmitted known signal is represented by s _u = ( _{su 1} , _{su 2} ,..., Su _{u, N} ) ^T. At this time, the decoded signal s _u ′ (t) = (s _{u, 1} ′ (t), _{su, 2} ′ (t),..., _Su, obtained from the equation (10) for the u-th subcarrier _{. N} ′ (t)) ^T is

と表せる。ここで、（１２）式のψ_ｕ,ｋ(t)（１≦ｋ≦Ｎ）は（１１）式のψ_k(t)に対応する。また、（g_１(t) , g_２(t) , ・・・ , g_Ｎ(t)）^ＴはＨ_ｕ ^−１ｎ_ｕ’(t)による熱雑音を表すが、重みＨ_ｕ ^−１が乗算されているため、それぞれ異なる大きさを持つ。既知信号ｓ_ｕの対応する要素で、ｓ_ｕ’ベクトルの各要素を割り算することで、各信号系列の位相回転量の推定値ψ’_ｕ,１(t)〜ψ’_ｕ,Ｎ(t)（Ｎ×１ベクトル）は以下のように表せる。

It can be expressed. Here, ψ _{u, k} (t) (1 ≦ k ≦ N) in the equation (12) corresponds to ψ _k (t) in the equation (11). Also, (g ₁ (t), g ₂ (t),..., G _N (t)) ^T represents thermal noise due to H _u ⁻¹ n _u ′ (t), but the weight H _u ⁻¹ is Since they are multiplied, they have different sizes. By dividing each element of the s _u ′ vector by the corresponding element of the known signal s _u , an estimated value ψ ′ _{u, 1} (t) to ψ ′ _{u, N} (t) of the phase rotation amount of each signal series (N × 1 vector) can be expressed as follows.

またさらに既知信号ｓ_ｕ,１〜ｓ_ｕ,Ｎの大きさをそれぞれ１と仮定すれば上式は、

Furthermore, assuming that the magnitudes of the known signals su _{, 1 to} su _{, N} are 1 respectively,

と各信号系列での位相回転量を推定することができる。なお、既知信号が送信される周波数帯は一つである必要はなく、複数のサブキャリアから得られる既知信号から各信号系列の位相回転量を推定することができる。

And the amount of phase rotation in each signal series can be estimated. Note that the frequency band in which the known signal is transmitted need not be one, and the phase rotation amount of each signal series can be estimated from the known signal obtained from a plurality of subcarriers.

The phase rotation compensation is realized by using the estimated value ψ ′ _{u, 1} (t) to ψ ′ _{u, N} (t) of the phase rotation obtained as described above for each signal series (claim 1). In general, there are few frequency bands in which known signals can be transmitted, and the noises g _{1 to} g _N included therein are considered to be non-negligible due to the effects of fading, inter-signal interference, and the like. If they are not synchronized, they cannot be removed and the phase rotation cannot be fully compensated. Here, using a received signal of a known signal in one or a plurality of frequency bands, and assuming that the phase rotation amount for the estimated k-th signal sequence is Θ _k ,

と表すことができる。ここで、θ_n,1(t)〜θ_n,N(t)は位相回転量の推定誤差を表し、熱雑音g_１〜g_Ｎに対応する大きさを持つため信号系列においてその大きさは異なる。また、時間変動による位相変動θ(t) は全信号系列で共通して用いることができ、伝達係数行列推定時の受信タイミングの違いにより生じる初期位相誤差θ(T_１)〜θ(T_Ｎ)は各信号系列で推定する必要がある。

It can be expressed as. Here, θ _{n, 1} (t) to θ _{n, N} (t) represents an estimation error of the phase rotation amount, and has a magnitude corresponding to the thermal noise g _{1 to} g _N. Different. Further, the phase variation θ (t) due to the time variation can be commonly used in all signal sequences, and the initial phase errors θ (T ₁ ) to θ (T _N ) caused by the difference in reception timing at the time of estimating the transfer coefficient matrix. Must be estimated for each signal sequence.

  Therefore, the phase rotation correction amount calculation device 171 calculates the time fluctuation correction phase Γ (t) as follows in order to remove the time fluctuation of the phase due to θ (t).

ここで、Ｗ_ｋは各信号系列の位相回転量推定値の確からしさに対応する重みづけであって、ΣＷ_ｋは１であり、（１５）式の推定誤差θ_ｎ,１(t)〜θ_ｎ,Ｎ(t)の影響を軽減するように設定される。（１６）式に（１５）式を代入してΓ(t)を書き直すと

Here, W _k is a weight corresponding to the certainty of the phase rotation amount estimation value of each signal series, ΣW _k is 1, and the estimation error θ _{n, 1} (t) to θ in equation (15) It is set so as to reduce the influence of _{n, N} (t). Substituting equation (15) into equation (16) and rewriting Γ (t)

と表すことができる。θ_ｎ,ｋ(t)は雑音成分であるため、足しあわされることでその大きさは小さくなっており、位相回転の時間変動について精度よく補正を行うことができる。復号信号ｓ_ｉ ^’ (t)は共通位相回転量補正装置１８０において、まずこの時間変動補正位相Γ(t)により補正を受ける。

It can be expressed as. Since θ _{n, k} (t) is a noise component, its magnitude is reduced by adding it, and it is possible to accurately correct the temporal variation of the phase rotation. The decoded signal s _i ^′ (t) is first corrected in the common phase rotation amount correction device 180 by this time variation correction phase Γ (t).

Next, the phase rotation correction amount calculation device 171 calculates the inter-signal sequence phase correction amounts γ _{1 to} γ _N for compensating for the phase shift occurring in each signal sequence as follows.

Ｔ_st(t)は積分を開始する受信タイミング、Ｔ_fin(t)は積分を終了する受信タイミングであり、w(t,k)はｔ番目の受信タイミングの信号系列間位相補正量γを求めるために用いるk番目の受信タイミングの位相変化量に乗算する重みであり、kについて積分すると１になるように決定され、（１５）式の推定誤差θ_ｎ,１(t)〜θ_ｎ,Ｎ(t)の影響を軽減するように設定される。（１８）式に（１５）式を代入してγ₁〜γ_Nを書き直すと、

T _{st (t)} is a reception timing at which integration is started, T _{fin (t)} is a reception timing at which integration is completed, and w (t, k) is a phase correction amount γ between signal sequences at the t-th reception timing. Is a weight for multiplying the phase change amount of the k-th reception timing used for the purpose, and is determined to be 1 when integrated with respect to k, and the estimation errors θ _{n, 1} (t) to θ _{n, N of} equation (15) are determined. It is set to reduce the effect of (t). Substituting equation (15) into equation (18) and rewriting γ ₁ to γ _N ,

と表すことができ、Γ（ｔ）に対する各信号系列の位相補正値が示されている。このように複数の受信タイミングのデータを利用することで、推定誤差の影響はより低減される。ここで得られたγ₁〜γ_Nを信号系列間位相差補正装置１８０−１〜１８０−Ｎにおいて各信号系列の復号信号に補正を行うことで位相回転による伝送品質劣化を低減することができる。

The phase correction value of each signal series with respect to Γ (t) is shown. In this way, the influence of the estimation error is further reduced by using data at a plurality of reception timings. By correcting the obtained γ ₁ to γ _N to the decoded signal of each signal sequence in the signal sequence phase difference correction devices 180-1 to 180-N, it is possible to reduce the transmission quality deterioration due to the phase rotation. .

From the above, the phase rotation correction in the common phase rotation amount correction device 180 and the signal sequence phase difference correction devices 180-1 to 180-N is performed as follows. The decoded signal s _i ′ (t) obtained by the equation (10) is obtained by substituting the equation (11) into the equation (12).

と書き直すことができる。以下、位相回転の補正に着目するため、熱雑音の項を省略し、

Can be rewritten. In the following, in order to focus on correction of phase rotation, the thermal noise term is omitted,

と表すことにする。共通位相回転量補正装置１８０では、まずｅｘｐ(−ｊΓ(t))により時間変動成分を補正する。すなわち、（２１）式にｅｘｐ(−ｊΓ(t))を乗算し（１７）式を代入して得られる一次補正復号信号をｓ_ｉ ^’’とすると、

It will be expressed as In the common phase rotation amount correction apparatus 180, first, the time variation component is corrected by exp (−jΓ (t)). That is, if the primary corrected decoded signal obtained by multiplying exp (−jΓ (t)) by equation (21) and substituting equation (17) is s _i ^″ ,

と表すことができ、各信号系列において時間変動による位相変化はなくなり、一定した位相のずれが得られるようになる。さらに、信号系列間位相差補正装置１８０−１〜１８０−Ｎでは、各信号系列にｅｘｐ(ｊγ₁(t))〜ｅｘｐ(ｊγ_N(t))を乗算し、これら位相のオフセットを除去する。すなわち、（２２）式の各成分にそれぞれｅｘｐ(ｊγ₁(t))、・・・、ｅｘｐ(ｊγ_N(t))を乗算して得られる二次補正復号信号をｓ_ｉ ^’’’とすると、

In each signal series, there is no phase change due to time variation, and a constant phase shift can be obtained. Further, inter-signal sequence phase difference correction apparatuses 180-1 to 180-N multiply each signal sequence by exp (jγ ₁ (t)) to exp (jγ _N (t)) to remove the offset of these phases. . That is, the secondary correction decoded signal obtained by multiplying each component of the expression (22) by exp (jγ ₁ (t)),..., Exp (jγ _N (t)) is expressed as s _i ^{″ ″.} Then

と表すことができ、位相回転は位相雑音θ_ｎ,ｋ(t)のみから構成されるものとなっている。また、多数の位相雑音を足し合わせていることから、位相誤差の著しい軽減が期待できる。（請求項２、３、６）

The phase rotation is composed only of the phase noise θ _{n, k} (t). In addition, since a large number of phase noises are added, a significant reduction in phase error can be expected. (Claims 2, 3, 6)

  Further, in the phase rotation correction amount calculation device 171, the time fluctuation correction phase Γ (t) can be calculated as follows in order to remove the time fluctuation of the phase due to θ (t).

ここで、ω_ｌ,ｔは受信タイミングｔの復号信号に用いる、l番目の受信タイミングの位相回転量推定値に乗算する重みづけであり、ω_ｌ,ｔをlについて受信タイミングＴ_st(t)からＴ_fin(t)まで積分すると１となるように決定される。このように制御することで、位相雑音の加算回数を増やすことができ、より位相回転量の推定誤差を低減することができる。（請求項２、４）

Here, ω _{l, t} is a weight used to multiply the estimated value of the phase rotation amount at the l-th reception timing used for the decoded signal at the reception timing t, and ω _{l, t} is the reception timing T _{st (t)} for l. _Is integrated to T _{fin (t)} to be 1. By controlling in this way, the number of additions of phase noise can be increased, and the estimation error of the phase rotation amount can be further reduced. (Claims 2 and 4)

  Further, in the phase rotation correction amount calculation device 171, the time fluctuation correction phase Γ (t) is assumed by some function in order to remove the time fluctuation of the phase due to θ (t), and is determined by fitting to that function. be able to. For example, the following linear function is assumed.

得られる位相回転量Θ_k(t)とその信頼度を考慮し、上式との誤差が最小となるようなａとｂを算出し、ここで得られたΓ(t)を共通位相回転量補正装置１８０において用いることができる。（請求項５）

Considering the obtained phase rotation amount Θ _k (t) and its reliability, a and b are calculated so that the error from the above equation is minimized, and Γ (t) obtained here is used as the common phase rotation amount. It can be used in the correction device 180. (Claim 5)

Further, when it is necessary to obtain the transfer coefficient matrix H _i for determining transmission weight, for example, γ ₁ (t) to γ _N (t) or Σγ ₁ (t) to Σγ _N (t) It can be obtained by multiplying the primary estimated transfer coefficient matrix H _i ^′ by an N × N diagonal matrix having angular components. (Claim 7)

Further, when it is necessary to determine the transfer coefficient matrix H _i , for example, for determining the transmission weight, the time fluctuation correction phase Γ (t) of the time fluctuation correction phase and the transfer coefficient matrix estimation timings T _{1 to} T _{N are used.}

と表すことができるＮ×Ｎの対角行列を一次推定伝達係数行列Ｈ_ｉ ^’に乗算することで得ることができる。（請求項８）

Can be obtained by multiplying the primary estimated transfer coefficient matrix H _i ^′ by an N × N diagonal matrix. (Claim 8)

In the above control method, it is assumed that the time variation of the phase is θ (t) and is the same for all signal sequences. However, in actuality, the value may be different for each signal sequence due to a slight shift in synchronization. It is possible. However, assuming that the time-varying component of the k-th signal sequence is θ _k (t),

が成り立てば、この変化はγ₁(t)〜γ_N(t)によって補償することができ、同様に用いることができる。

If this holds, this change can be compensated by γ ₁ (t) to γ _N (t) and can be used similarly.

In the above control method, it is assumed that the known signal S ₀ for estimating the transfer coefficient matrix is a diagonal matrix. However, when S ₀ is not a diagonal matrix, matrixes in which column vectors are orthogonal to each other are used. Is based on the phase shift of the primary estimated transfer coefficient matrix, for example, by using the repetitive part of the known signal for determining the OFDM acquisition point or repeating the actual transmission signal as the guard interval of the OFDM signal. By correcting from the result of the rough estimation of the phase, similarly, the rotation of the phase can be accurately corrected by the above control method.
In addition, the estimation of the amount of phase rotation has been described below using equation (16) using the phase θ of the signal point, but the processing such as averaging using the weight corresponds to the phase rotation represented by exp (−jφ) or the like. Operators can be used as well.
In addition, when each of a plurality of transmission devices has a plurality of transmission elements, even when an independent signal sequence is transmitted from each antenna or transmission by directivity control is performed, from each signal sequence If a known signal is transmitted in advance and the transfer coefficient is known, the same processing can be performed.

  Hereinafter, experimental results of the present invention will be described with reference to FIGS. In this experiment, IEEE 802.11a (IEEE, “Part 11: Wireless LAN Medium Access Control (MAC) and IEEE 802.11a) between four transmitting devices each having one antenna and a receiving device having four antennas. Physical Layer (PHY) specifications: High-speed Physical Layer in the 5 GHz Band ", IEEE 802.11a-1999, Dec. 1999.) The four independent signal sequences were transmitted by the spatial multiplexing division method and the orthogonal frequency division multiplexing method.

  A block diagram of the four signal sequences used for transmission is shown in FIG. The transmission signal is composed of a sampling synchronization known signal, four transmission coefficient matrix estimation known signals, and 50 data signals, each of which was transmitted at the timing shown in the figure. The average reception level at each receiving antenna was about 30 dB. Of the 64 frequency bands obtained in each of the four signal sequences, the central 53 frequency band is used, and the 26th frequency band as the center is not used. Then, it is assumed that known signals for phase synchronization are put in the sixth, twentieth, thirty-fourth, and forty-eighth frequency bands, and phase rotation compensation of four signal sequences is performed based on these pieces of information.

  FIG. 3 is a graph showing the result of estimating the phase rotation amount by the conventional method. That is, in each signal series, the known signal in each frequency band is weighted according to the reception level, and the phase rotation amount is estimated. The horizontal axis represents the received symbol timing, and the vertical axis represents the amount of phase rotation. FIG. 3 shows that the phase rotation amount estimated by the symbol timing varies and the phase rotation compensation is not performed with high accuracy.

On the other hand, FIG. 4 shows the result of estimating the amount of phase rotation by the method and apparatus of claims 2 and 4. In equation (18), T _{fin (t) is set} to t + 2, T _{st (t) is set} to t−2, w (t, k) is set to 1/5, and γ _i (t) at t = 1, 2 Γ _i (3) is used, and γ _i (47) is used as γ _i (t) at t = 49,50. Further, T _{fin (t)} in equation (24) is t + 1, T _{st (t)} is t−1, ω _{t−1, t} = 1/3, ω _{t, t} = 2/3, ω _{t + 1, t} = 2/3, and W _k is set as the reciprocal of the norm value of the row component of the inverse matrix of the transfer coefficient matrix corresponding to the k-th signal sequence. FIG. 4 shows that the method and apparatus of claims 2 and 4 can reduce an error due to noise and accurately obtain the phase rotation amount.

  According to the present invention, it is possible to reduce deterioration in transmission quality due to phase rotation in communication using space division multiplexing.

It is a block diagram which shows the structure of the receiver for spatial multiplexing transmission in embodiment of this invention. It is a block diagram of four signal series used for transmission in experiment. It is a graph which shows the result of having estimated the amount of phase rotation by the conventional method using an experimental device. It is a graph which shows the result of having estimated the amount of phase rotation by this invention using an experimental device. It is a block diagram which shows the structure of the receiver using the conventional orthogonal wave frequency division multiplexing.

Explanation of symbols

100-1, 100-2 to 100-M ... Antenna element 110-1, 110-2 to 110-M ... AGC
120-1, 120-2 to 120-M ... A / D conversion device 130 ... Signal acquisition point determination device 140 ... FFT device 150 ... Transfer coefficient matrix primary estimation device 160 ... Spatial division multiplexing decoding device 170-1 to 170-N ... known signal phase detector 171 ... phase rotation correction amount calculation device 180 ... common phase rotation amount correction device 181-1 to 181-M ... phase difference correction device between signal sequences 190 ... subcarrier decoding device 900 ... antenna element 910 ... AGC
920 ... A / D conversion device 930 ... Signal acquisition point determination device 940 ... FFT device 950 ... Transfer coefficient estimation device 960 ... Transfer coefficient equivalent device 970 ... Known signal phase detection device 980 ... Phase rotation amount correction device 990 ... Subcarrier decoding device

Claims (8)

The receiver has multiple antenna elements and uses orthogonal frequency division multiplexing to compensate for phase rotation of signal points when receiving and decoding signal sequences transmitted from multiple transmitters at the same time and frequency. A way to
Receiving a known signal transmitted at a plurality of transmission timings from a transmission device, calculating a primary estimated transmission coefficient matrix for estimating a transmission coefficient matrix including a phase shift between the transmission device and the reception device, and transmitting the primary estimation transmission The received signal is decoded using a coefficient matrix, and all signal sequences are obtained from known signals used in a specific frequency band of the orthogonal frequency division multiplexing received at the same time or a plurality of times included in the decoded signal. A receiving method for spatial multiplexing transmission, comprising a step of calculating an instantaneous phase correction amount and correcting a time variation of phase rotation of all signal sequences using the instantaneous phase correction amount. The reception method for spatial multiplexing transmission according to claim 1,
When the simultaneously received signal sequences are not completely synchronized, the phase rotation amount of each signal sequence is calculated from known signals included in the reception signals at a plurality of reception timings, and the phase rotation amount of each signal sequence is calculated. A receiving method for spatial multiplexing transmission, comprising: calculating a difference from the instantaneous phase correction amount as a phase rotation correction amount and performing phase rotation correction of each signal series using the phase rotation correction amount. Spatial multiplex transmission receiver for decoding transmission signals of N signal sequences transmitted at the same time using orthogonal wave frequency division multiplexing with M received signals obtained at the same time from M antenna elements In
An A / D converter connected to each of the antenna elements for analog / digital conversion;
A signal acquisition point determination device that is connected to the A / D conversion device and determines an acquisition point of an OFDM signal of an orthogonal wave frequency division multiplexing system from a known signal for sampling synchronization of the obtained data;
An FFT device that performs Fourier transform on the OFDM signal and calculates a signal in each frequency band;
A transfer coefficient matrix primary estimation device that is connected to the FFT device and calculates a primary estimation transfer coefficient matrix from an OFDM known signal for estimating a transfer coefficient matrix included in the OFDM signal;
Connected to the transmission coefficient matrix primary estimation device, decodes M OFDM signals received at the same time using the primary estimation transmission coefficient matrix, and outputs a decoding result of a known signal included in each signal sequence And a space division multiplex decoding device that outputs a decoding result of the desired data signal;
Calculates the amount of phase rotation of each signal sequence at a predetermined time from the decoding result of the known signal of each signal sequence provided for each signal sequence and input from the spatial division multiplexing decoder, and outputs the result A known signal phase detection device,
It is connected to the known signal phase detection device and calculates and outputs the phase rotation amount of the entire signal sequence from the phase rotation amount at a predetermined time of each input signal sequence. Phase rotation correction that calculates the phase rotation amount of each signal sequence by averaging the phase rotation amount at the symbol timing and outputs the difference of the calculated phase rotation amount of each signal sequence from the phase rotation amount of all the signal sequences A quantity calculation device;
The phase rotation of the entire signal sequence at a time corresponding to the decoding result of the desired data signal input from the phase rotation correction amount calculating device, with respect to the decoding result of the desired data signal input from the spatial division multiplexing decoding device A common phase rotation amount correction device that performs signal point correction based on the amount and outputs a primary correction decoding result;
The difference between the phase rotation amount of each signal sequence and the phase rotation amount of all the signal sequences input from the phase rotation correction amount calculation device with respect to the primary correction decoding result input from the common phase rotation amount correction device. A signal point phase difference correction device that performs signal point correction based on the above and outputs a secondary correction decoding result;
A subcarrier decoding apparatus that is connected to the inter-signal-sequence phase difference correction apparatus and calculates desired bit data from signal points of the input secondary correction decoding result;
A receiver for spatial multiplexing transmission, comprising:   The known signal phase detection device calculates an estimated value of a temporal change in phase from the phase rotation amounts of all signal sequences determined from the phase rotation amounts of a plurality of input symbol timings, and temporally rotates the phase rotation amount. 4. The spatial multiplexing transmission receiver according to claim 3, wherein the phase rotation amount of all signal sequences at a predetermined symbol timing is determined so as not to change, and is output to the common phase rotation amount correction device. .   The known signal phase detection device performs fitting to a given function from the phase rotation amounts of all signal sequences determined from the phase rotation amounts of a plurality of input symbol timings, and calculates a time change amount of the phase, 4. The spatial multiplexing transmission receiver according to claim 3, wherein the phase rotation amount of all signal sequences at a predetermined symbol timing is determined so as not to cause phase rotation, and is output to the common phase rotation amount correction device. . The space division multiplex decoding apparatus performs decoding on M OFDM signals received at the same time using the primary estimated transfer coefficient matrix, decodes a known signal included in each signal sequence, and the primary estimated transfer Output the probability of the known signal given as an estimated value to the coefficient matrix to the known signal phase detection device, and output the decoding result of the desired data signal to the common phase rotation amount correction device,
The known signal phase detection device weights or selects the decoding result of the known signal based on the decoding result of the input known signal and its certainty, determines the phase rotation amount of each signal series, and calculates the phase rotation correction amount Output the results and certainty to the device,
The phase rotation correction amount calculation device performs weighting or selection based on the phase rotation amount of each signal sequence input from the known signal phase detection device and its certainty, and calculates the phase rotation amount of all signal sequences. Output to the common phase rotation amount correction device, and weight or select the phase rotation amount at a plurality of symbol timings of each input signal sequence on the basis of the input probability. 5. The spatial multiplexing transmission receiver according to claim 3, wherein a difference from the phase rotation amount is calculated and output to the inter-signal sequence phase difference correction apparatus.   When using a transfer coefficient matrix representing an actual propagation environment including the case of obtaining a transmission weight that is optimal for the propagation environment, each signal sequence obtained by the phase rotation correction amount calculation device for the primary estimated transfer coefficient matrix 6. The spatial multiplexing transmission according to claim 3, wherein a transmission coefficient matrix obtained by performing correction using a difference between the phase rotation amount of the signal sequence and the phase rotation amount of the entire signal sequence is used. Receiving device. When using a transfer coefficient matrix representing an actual propagation environment including the case of obtaining a transmission weight that is optimal for the propagation environment, all signal sequences obtained by the phase rotation correction amount calculation device for the primary estimated transfer coefficient matrix Using a transfer coefficient matrix obtained by performing correction using a temporal phase change amount obtained from the phase rotation amount of the signal and a phase change amount obtained from the symbol timing obtained by calculating the primary estimated transfer coefficient matrix. The receiving device for spatial multiplexing transmission according to any one of claims 3 to 5.

Images