JP2010193350A - Communication apparatus and communication system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a communication apparatus and the like in which an error rate characteristic is enhanced by appropriately selecting such a method that an instantaneous signal-to-interference and noise power ratio (SINR) is improved in detecting the separation of a desired signal in MIMO-SDM. <P>SOLUTION: A communication apparatus comprises: a propagation path estimating section for receiving a plurality of spatially divided and multiplexed streams at a plurality of receiving antennas to estimate a propagation path matrix and noise power; a weight calculating section for calculating a weight matrix on the basis of the propagation path matrix and noise power; a synthesization section for synthesizing a signal for each stream from the reception signals received by the receiving antennas and the weight matrix; a system selection section for selecting either the reception signal or the synthesized signal; a demapping section for calculating the likelihood information of the selected signal on the basis of the propagation path matrix or on the basis of the propagation path matrix and the weight matrix; and a decoding section for performing error-correcting decoding on the basis of the likelihood information. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a communication apparatus that receives a plurality of space-division multiplexed streams with a plurality of receiving antennas.

  In a multiple-input multiple-output (MIMO) system that uses multiple antennas for transmission and reception, an SDM (Space Division Multiplexing) system that improves the transmission efficiency by sending independent signals from each transmit antenna Various studies have been conducted (for example, see Non-Patent Document 1 and Non-Patent Document 2).

  Since MIMO-SDM receives a signal in which a plurality of signals are superimposed, it is necessary to separately detect a desired signal. As the method, there are known separation methods based on MLD (Maximum Likelihood Detection) based on maximum likelihood estimation and MMSE (Minimum Mean Square Error) standard which is linear processing. The separation method by MMSE can greatly reduce the amount of calculation compared with MLD.

  On the other hand, in a cellular system using an OFDM (Orthogonal Frequency Division Multiplex) scheme, when there is co-channel interference occurring between cells, the likelihood considering the occurrence probability of interference signals during error correction decoding There has been proposed a technique for suppressing interference by obtaining and applying a degree (see, for example, Non-Patent Document 3).

  Further, when a desired signal is separated in MIMO-SDM, signals other than the desired signal (demodulation target signal) are regarded as interference signals, and a likelihood that considers the occurrence probability distribution of the interference signals at the time of error correction decoding is obtained. A method for suppressing the interference signal and improving the signal separation accuracy has been proposed (see, for example, Non-Patent Document 4).

Yoshio KARASAWA, "MIMO Propagation Channel Modeling", IEICE Trans. Commun., Vol.E88-B, no.5, pp.1829-1842, May 2005 Takeo OHGANE et al., "Applications of Space Division Multiplexing and Those Performance in a MIMO Channel", IEICE Trans. Commun., Vol.E88-B, no.5, pp. 1843-1851, May 2005 Hira et al., "Soft decision likelihood information generation method for COFDM systems under co-channel interference", IEICE (B), Vol. J91-B, No. 5, pp. 575-584, May 2008 Yoshihara et al., "Application of interference suppression method by likelihood operation in error correction decoding to correlated channel MIMO-SDM system", 2008 Sci. Sogaku Univ., B-8-2, September 2008

  However, in the conventional signal separation method based on the MMSE standard, when the independence of a plurality of propagation paths set between transmitting and receiving antennas is lost and there is a correlation, the signal separation accuracy is remarkably deteriorated and the error rate characteristic is deteriorated. was there.

  Further, when separating a desired signal in MIMO-SDM, a method of obtaining a likelihood considering the occurrence probability distribution of an interference signal at the time of error correction decoding and applying a signal other than the desired signal as an interference signal is as follows. The error rate characteristic can be realized almost constant regardless of the correlation level of the propagation path, but the error rate characteristic is inferior in the case of the propagation path having a small correlation as compared with the signal separation method based on the MMSE standard. was there.

  Furthermore, when the above two signal separation / detection methods are used, which method can realize better error rate characteristics depends on the magnitude of the correlation of the propagation path, the S / N ratio of the received signal ( Signal-to-noise power ratio), S / N ratio after MMSE synthesis, and interference suppression effect in error correction decoding using likelihood considering interference signal occurrence probability distribution. It was not possible to judge.

  In view of the above-described problems, an object of the present invention is to improve instantaneous SINR (Signal to Interference plus Noise power Ratio) in separation detection of a desired signal in MIMO-SDM. It is possible to provide a communication apparatus or the like that can appropriately select a method and has improved error rate characteristics.

  In view of the above-described problems, the communication device of the present invention is a communication device that receives a plurality of spatially multiplexed streams with a plurality of reception antennas, a propagation path estimation unit that estimates a propagation path matrix and noise power, A weight calculating unit that calculates a weight matrix based on the propagation path matrix and noise power, a combining unit that combines a received signal received by the receiving antenna and a signal for each stream from the weight matrix, the received signal and the combining A method selection unit that selects one of the selected signals, a demapping unit that calculates likelihood information of the selected signal based on the propagation path matrix or the propagation path matrix and the weight matrix, and the likelihood And a decoding unit that performs error correction decoding based on the degree information.

  In the communication apparatus according to the present invention, the weight calculation unit calculates an MMSE weight matrix for separating the received signal into the plurality of streams based on a minimum mean square error criterion.

  Further, in the communication device according to the present invention, the demapping unit may be configured to generate a residual interference component after the synthesis for each stream when a signal for each stream synthesized by the scheme selection unit is selected. The likelihood information is calculated.

  In the communication apparatus of the present invention, when the received signal is selected by the scheme selecting unit, the demapping unit is a stream for which likelihood information is calculated among a plurality of streams superimposed on the received signal. The likelihood information is generated based on an occurrence probability distribution related to another stream that is an interference component excluding.

  Further, the communication apparatus of the present invention calculates a first signal-to-interference and noise power ratio for each combined stream based on the weight matrix, the propagation path matrix, and noise power, and the propagation path matrix and An SINR calculation unit that calculates a second signal-to-interference and noise power ratio for each stream of the received signal based on noise power, and the method selection unit includes the first signal-to-interference and noise power ratio Is greater than the second signal-to-interference and noise power ratio, the combined signal for each stream is selected, and the second signal-to-interference and noise power ratio is the first signal-to-interference and noise power. The received signal is selected if the power ratio is greater than the power ratio, and the combined signal for each stream is selected if the first signal-to-interference and noise power ratio is equal to the second signal-to-interference and noise power ratio. And selects either the serial reception signal.

  In the communication apparatus of the present invention, the SINR calculation unit calculates the first and second signal-to-interference and noise power ratios using interference signal power suppressed by the error correction decoding. To do.

  The communication system of the present invention is a communication system including a transmitter that transmits a plurality of streams by space-division multiplexing from a plurality of transmission antennas, and a receiver that receives signals transmitted from the transmitter by a plurality of reception antennas. The receiver includes a propagation path estimation unit that estimates a propagation path matrix and noise power, a weight calculation unit that calculates a weight matrix based on the propagation path matrix and noise power, and reception received by the receiving antenna. A synthesis unit that synthesizes a signal for each stream from the signal and the weight matrix, a method selection unit that selects one of the received signal and the synthesized signal, and the propagation path matrix or the propagation path matrix and the weight matrix. And a decoding unit that calculates likelihood information of the selected signal based on the likelihood information, and a decoding unit that performs error correction decoding based on the likelihood information. That.

  According to the present invention, it is possible to select a method in which instantaneous SINR (Signal to Interference plus Noise power Ratio) is improved in separation detection of a desired signal in MIMO-SDM, A communication apparatus and a communication system with improved error rate characteristics can be provided.

It is a figure for demonstrating the structure of the system in this embodiment. It is a figure for demonstrating the occurrence probability distribution in this embodiment. It is a figure for demonstrating the structure of the transmitter in this embodiment. It is a figure for demonstrating the structure of the receiver in this embodiment. It is a figure for demonstrating the structure of the MIMO isolation | separation part in this embodiment. It is a figure for demonstrating the characteristic in the communication system at the time of applying this invention. It is a figure for demonstrating the characteristic in the communication system at the time of applying this invention. It is a figure for demonstrating the characteristic in the communication system at the time of applying this invention.

  Communication that performs communication using MIMO-SDM (Multiple-Input Multiple-Output-Space Division Multiplexing) in the OFDM (Orthogonal Frequency Division Multiplex) scheme below An embodiment in which the present invention is applied to a system and a communication apparatus will be described.

  First, FIG. 1 is a diagram showing a configuration of a 2 × 2 MIMO-SDM system in which two transmission / reception antennas are arranged, and transmission signals of different data sequences (streams) are transmitted from the two transmission antennas.

  The transmission signal matrix is S, the propagation path matrix is H, the reception signal matrix is R, the thermal noise matrix is N, and the MMSE (linear processing is performed so that the power sum of interference and noise is minimized as the MIMO-SDM signal separation. When using a method based on (Minimum Mean Square Error), the MMSE weight matrix is W, and the output signal matrix after MMSE is S ′.

  First, a method of applying a technique for suppressing interference by obtaining and applying a likelihood considering an occurrence probability distribution of an interference signal at the time of error correction decoding to a residual interference component after MMSE (hereinafter referred to as a first method). Will be described.

ただし、Ａ_ij＝Ｗ_ijＨ_ij（ｉ＝１，２、ｊ＝１，２）である。 Each element of the output signal matrix S ′ after MMSE in the MIMO-SDM system shown in FIG.

However, A _ij = W _ij H _ij (i = 1, 2, j = 1, 2).

From Equation 1, it can be seen that A ₁₂ S ₂ remains as an interference component for the signal S ₁ ′ and A ₂₁ S ₁ remains as an interference component for the signal S ₂ ′ even after MMSE.

ここで、振幅±ｍを等確率でとる変調方式ＢＰＳＫ（Binary Phase Shift Keying）を例にとって説明する。このとき、信号Ｓ_１’に関する干渉成分Ａ_１２Ｓ_２は±ｍＡ_１２を等確率でとる。図２は、この様子を示したものである。したがって、信号Ｓ_１’に対する干渉成分の生起確率分布Ｐ_Ｉ（Ｓ_２）は数３で表現される。

ただし、δはディラックのデルタ関数である。 For residual interference component after the MMSE, using a Gaussian distribution P _N of the probability distribution P _I and a noise component, and calculates the likelihood P (u) used at the time of error correction decoding based on the number 2.

Here, a modulation method BPSK (Binary Phase Shift Keying) that takes amplitude ± m with equal probability will be described as an example. At this time, the interference component A ₁₂ S ₂ related to the signal S ₁ ′ takes ± mA ₁₂ with an equal probability. FIG. 2 shows this state. Therefore, the occurrence probability distribution P _I (S ₂ ) of the interference component with respect to the signal S ₁ ′ is expressed by Equation 3.

Where δ is a Dirac delta function.

以上より、信号Ｓ_１’とＳ_２’に関する尤度Ｐ_１（ｕ）とＰ_２（ｕ）は数５で与えられる。

ただし、σ_１ ^２とσ_２ ^２は信号Ｓ_１’とＳ_２’に関する雑音成分電力を意味し、数６で表せる。

なお、σ_ｎは熱雑音の標準偏差である。 Further, by the same discussion, the occurrence probability distribution P _I (S ₁ ) of the interference component A ₂₁ S ₁ related to the signal S ₂ ′ is expressed by Equation 4.

From the above, the likelihoods P ₁ (u) and P ₂ (u) relating to the signals S ₁ ′ and S ₂ ′ are given by Equation 5.

However, σ ₁ ² and σ ₂ ² mean noise component powers related to the signals S ₁ ′ and S ₂ ′, and can be expressed by Expression 6.

Note that σ _n is the standard deviation of thermal noise.

  By introducing the likelihood calculated based on Equation 5 at the time of error correction decoding, the residual interference component after MMSE can be suppressed, and the signal separation accuracy by MMSE can be further improved. In addition, the likelihood can be calculated in the same manner for modulation schemes other than BPSK.

  Next, a technique for suppressing interference by obtaining and applying likelihood considering the occurrence probability distribution of interference signals at the time of error correction decoding without performing MMSE is superimposed on the receiving antenna in MIMO-SDM. A method (hereinafter referred to as a second method) applied to an interference signal that is not subject to demodulation among a plurality of signals will be described.

  In the case of a correlated channel, in the signal separation method based on MMSE, the signal separation accuracy deteriorates as the correlation of the channel increases. Therefore, in the case of such a propagation path, without performing signal separation by MMSE, it is assumed that one of the signals arriving superimposed on the receiving antenna is a desired signal and the other is an interference signal, and the occurrence probability distribution of the interference signal is The interference signal is suppressed by using the considered likelihood at the time of error correction decoding.

At this time, each element of the received signal matrix R can be expressed by Equation 7.

ただし、σ_ｎは熱雑音の標準偏差である。 In Equation 7, H ₂₁ S ₂ regarding the signal R _{1 and} H ₁₂ S ₁ regarding the signal R ₂ are regarded as interference components, and by calculating the likelihood in consideration of the occurrence probability distribution, the interference signal is suppressed. Extract the signal. Based on the same discussion as above, the likelihoods P ₁ ′ (u) and P ₂ ′ (u) relating to the signals R ₁ and R ₂ when the modulation scheme is BPSK are given by Equation 8.

Where σ _n is the standard deviation of thermal noise.

  In the present embodiment, the above-described first method and second method are selected by selecting a method in which instantaneous SINR (Signal to Interference plus Noise power Ratio) is good in subcarrier units. Switch.

まず、ＭＭＳＥ後の残留干渉成分を抑圧する第１の方式における瞬時ＳＩＮＲを算出する。ＭＭＳＥ後の各ストリームの瞬時ＳＩＮＲγ_１とγ_２は、数１及び数９より数１０で表せる。

Assuming that the desired signal power is P _D , the interference signal power is P _I , and the noise power is P _N , γ can be expressed by Equation 9 when the true SINR is γ.

First, the instantaneous SINR in the first method for suppressing the residual interference component after MMSE is calculated. The instantaneous SINRs γ ₁ and γ ₂ of each stream after MMSE can be expressed by Equation 10 from Equation ₁ and Equation 9.

However, the interference signal powers | A ₁₂ | ² and | A ₂₁ | ² in Equation ¹⁰ are the power before suppression of the residual interference component, and in order to obtain the SINR directly related to the bit error, interference after suppression of the residual interference component It is necessary to calculate the signal power. Therefore, introducing an interference suppression amount _{T h} based on the likelihood applied, an interference signal power after interference suppression _| A ¹² _| 2 / _{T h} and _| A ²¹ | instantaneous SINRganma ₁ as 2 / _{T h} 'and gamma _2' Is calculated based on Equation (11).

By the same argument, the instantaneous SINRγ _{1 "and} gamma _2" of each stream in the second method of suppressing interference without MMSE, using the interference suppression amount T _h based on the number 7 and number 9 the likelihood applied The calculation is performed based on equation (12).

Γ _i ′ and γ _i ″ (i = 1, 2) are calculated based on Equations 11 and 12, and when γ _i ′> γ _i ″, the first method for suppressing the residual interference component after MMSE is: When γ _i ′ <γ _i ”, the second scheme for suppressing interference without performing MMSE is selected for each subcarrier.

As described above, instantaneous SINRs γ ₁ ′, γ ₂ ′, γ ₁ ″ and γ ₂ ″ after interference suppression in each method are calculated, and these are compared to obtain the first method and the second method. Therefore, it is possible to appropriately select a method with good error rate characteristics.

  Next, the configuration of the transmitter and the receiver will be described. FIG. 3 is a block diagram showing the configuration of the transmitter in this embodiment.

  Transmitter 300 includes coding sections 301a and 301b, modulation sections 302a and 302b, pilot generation section 303, pilot multiplexing sections 304a and 304b, frequency time conversion sections 305a and 305b, and GI insertion sections 306a and 306b. , Wireless transmission units 307a and 307b and transmission antenna units 308a and 308b.

  The encoding unit 301a and the encoding unit 301b each receive first transmission data and second transmission data, which are two different data sequences, or a data sequence obtained by dividing one data sequence into two sets, respectively. Is subjected to error correction encoding and encoded data is output.

Modulation section 302a and modulation section 302b modulate the encoded data output from encoding section 301a and encoding section 301b, respectively, and output modulated data symbols.
Pilot generation section 303 generates two sequences of pilot symbols (first transmission pilot symbol and second transmission pilot symbol) that are orthogonal to each other, and outputs them to pilot multiplexing section 304a and pilot multiplexing section 304b, respectively. Two series of pilot symbols orthogonal to each other are transmitted by time division, that is, one based on a sequence generated such that one is transmitting a known signal and the other is no signal, and frequency division is transmitted by subcarriers that do not overlap each other. However, it is preferable to use those generated based on sequences generated by code sequences orthogonal to each other. However, the present invention is not limited to these.

  Pilot multiplexing section 304a and pilot multiplexing section 304b multiplex and output the two sequences of pilot symbols generated by pilot generation section 303 with modulated data symbols, respectively.

  The frequency-time conversion unit 305a and the frequency-time conversion unit 305b perform time-frequency conversion, for example, IFFT (Inverse Fast Fourier Transform), on the modulated data symbol in which the pilot symbols are multiplexed, and generate an OFDM time signal. .

  The GI insertion unit 306a and the GI insertion unit 306b respectively add a GI (Guard Interval) to the OFDM time signal and output it.

The radio transmission unit 307a and the radio transmission unit 307b convert the OFDM time signal to which the GI is added into a radio signal, and transmit the radio signal from the transmission antenna unit 308a and the transmission antenna unit 308b, respectively. Transmission signals _{S 1} in FIG. 1 from the transmitting antenna unit 308a, the transmission signal _{S 2} is a signal transmitted from the transmitting antenna section 308b.

  FIG. 4 is a block diagram showing the configuration of the receiver in this embodiment. Receiver 400 includes reception antenna units 401a and 401b, radio reception units 402a and 402b, GI removal units 403a and 403b, time frequency conversion units 404a and 404b, pilot separation units 405a and 405b, and a propagation path estimation unit. 406, a MIMO separation unit 407, and decoding units 408a and 408b.

  The radio reception unit 402a and the radio reception unit 402b receive the radio signal transmitted from the transmitter 300 through the reception antenna unit 401a and the reception antenna unit 401b, respectively, convert the radio signal into a baseband OFDM time signal, and output it.

  The GI removal unit 403a and the GI removal unit 403b each remove the GI from the OFDM time signal and output it.

  The time-frequency conversion unit 404a and the time-frequency conversion unit 404b respectively perform time-frequency conversion, for example, FFT (Fast Fourier Transform) on the OFDM time signal from which the GI has been removed, and output modulation symbols.

  Pilot demultiplexing section 405a demultiplexes the modulation symbols output from time-frequency conversion section 404a into first received pilot symbols and first data symbols, and outputs them. Pilot demultiplexing section 405b demultiplexes the modulation symbols output from time-frequency conversion section 404b into second received pilot symbols and second data symbols, and outputs them.

The propagation path estimation unit 406 estimates the propagation path between each transmission antenna of the transmitter 300 and each reception antenna of the receiver 400 using the first reception pilot symbol and the second reception pilot symbol. A propagation path matrix H and noise power σ _n ² are obtained and output as propagation path estimation results.

The propagation path matrix H and the noise power σ _n ² are estimated as follows, for example. Since the first transmission pilot symbol transmitted from the transmission antenna unit 308a of the transmitter and the second transmission pilot symbol transmitted from the transmission antenna unit 308b are orthogonal to each other, the first reception pilot symbol and the second transmission pilot symbol The received pilot symbols can be further separated into the results of receiving the first transmission pilot symbol and the second transmission pilot symbol.

Here, for each subcarrier, the complex number representation of the first transmission pilot symbol is x ₁ , the complex number representation of the second transmission pilot symbol is x ₂ , and the first transmission pilot symbol separated from the first reception pilot symbol. The complex number representation of the reception result is y ₁₁ , the complex number representation of the reception result of the first transmission pilot symbol separated from the second reception pilot symbol is y ₁₂ , and the second transmission pilot separated from the first reception pilot symbol. If the complex number representation of the symbol reception result is y ₂₁ , and the complex number representation of the reception result of the second transmission pilot symbol separated from the second reception pilot symbol is y ₂₂ , the channel matrix H is estimated based on Equation 13. Is done.

In addition, a random noise component is canceled by obtaining an average value of received pilot symbols (any one or more of y ₁₁ , y ₁₂ , y _{21, and} y ₂₂ ) during a period in which the propagation path can be regarded as not changing. Is subtracted from the original received pilot symbol before the average, the noise component contained in the original received pilot symbol can be extracted. By obtaining the power of this noise component, the noise power σ _n ² can be estimated.

  The MIMO separation unit 407, based on the propagation path estimation result estimated by the propagation path estimation unit 406, two (encoded) data in which the first data symbol and the second data symbol are multiplexed by MIMO. Separation detection is performed on the series, and each likelihood is output. Details of the MIMO separation unit 407 will be described later.

  Decoding section 408a and decoding section 408b perform error correction decoding based on the likelihood of the two encoded data sequences output from MIMO separation section 407, and output first received data and second received data, respectively. .

  Next, the configuration of the MIMO separation unit 407 will be described with reference to FIG. The weight calculation unit 501 calculates and outputs an MMSE weight matrix W (weight matrix W) for MIMO signal separation based on the MMSE criterion from the propagation path estimation result output from the propagation path estimation unit 406.

  The MMSE combining unit 502 performs MMSE combining of the first data symbol and the second data symbol output from the pilot separating unit 405a and the pilot separating unit 405b using the MMSE weight matrix W calculated by the weight calculating unit 501. The first post-MMSE symbol and the second post-MMSE symbol are output.

The SINR calculation unit 503 uses the propagation path estimation result (the propagation path matrix H and the noise power σ _n ² ) output from the propagation path estimation unit 406 and the MMSE weight matrix W calculated by the MMSE weight calculation unit 501 to determine the first The instantaneous SINR γ ₁ ′ and γ ₂ ′ of each stream when using the above method is calculated based on Equation 11, and the instantaneous SINR γ ₁ ″ and γ ₂ ″ of each stream when using the second method is calculated using Equation 12. Calculate and output based on Incidentally, the method for determining the interference suppression amount T _h in Equation 11 and Equation 12 will be described later.

Scheme selection unit 504a, the first data symbol and a first MMSE after symbol is input, instantaneous SINRganma ₁ for the first stream calculated in SINR calculating unit 503 'compares the gamma ₁ ", gamma _1'> In the case of γ ₁ ″, the first symbol after MMSE is selected and output, and in the case of γ ₁ ′ <γ ₁ ″, the first data symbol is selected and output. γ ₁ ′ = γ ₁ ″ In this case, either one may be selected. Here, the first post-MMSE symbol is selected.

Method selecting section 504b, the second data symbol and the second MMSE after symbol is input, instantaneous SINRganma ₂ for the second stream calculated in SINR calculating unit 503 'compares the gamma ₂ ", gamma _2'> In the case of γ ₂ ″, the second post-MMSE symbol is selected and output. In the case of γ ₂ ′ <γ ₂ ″, the second data symbol is selected and output. γ ₂ ′ = γ ₂ ″ In this case, either one may be selected. Here, the second post-MMSE symbol is selected.

Demapping section 505a, the symbol selected by the mode selection unit 504a is input, instantaneous SINRganma ₁ for the first stream calculated in SINR calculating section 503 'and gamma ₁ "compare, gamma _1'> gamma _1" In the case of, the likelihood of each bit of the encoded data sequence is calculated based on Equation 5 using the propagation path estimation result and the MMSE weight matrix W, and in the case of γ ₁ ′ <γ ₁ ″, the propagation path estimation result is used. Then, the likelihood of each bit of the encoded data sequence is calculated based on Equation 8, and is output as the first likelihood information. Note that when γ ₁ ′ = γ ₁ ″, the likelihood is calculated by either method. Here, it is assumed that the likelihood is calculated based on Equation 5 using the propagation path estimation result and the MMSE weight matrix W.

Demapping unit 505b includes symbol selected by the mode selection unit 504b is input, instantaneous SINRganma ₂ for the second stream calculated in SINR calculating section 503 'and gamma ₂ "compares, gamma _2'> gamma _2" In the case of, the likelihood of each bit of the encoded data sequence is calculated based on Equation 5 using the propagation path estimation result and the MMSE weight matrix W, and in the case of γ ₂ ′ <γ ₂ ″, the propagation path estimation result is used. Then, the likelihood of each bit of the encoded data sequence is calculated based on Equation 8 and is output as second likelihood information. Note that, when γ ₂ ′ = γ ₂ ″, the likelihood is calculated by either method. Here, it is assumed that the likelihood is calculated based on Equation 5 using the propagation path estimation result and the MMSE weight matrix W.

6, BER against acquired by computer simulation interference suppressing amount T _h: is a graph of (Bit Error Ratio BER) characteristics.

In the computer simulation, it is assumed that the sampling period is T and the OFDM symbol period is T _S (= 80T), and symbol synchronization and frequency synchronization between the transceivers are perfect. Also, all delayed waves are assumed to be within the GI. Assuming LOS (Line Of Sight) environment and NLOS (Non-Line Of Sight) environment as propagation environment, Rice factor K representing the power ratio of direct wave and reflected wave in LOS environment is set to 3 dB In addition, the worst environment in which direct waves arrive at each receiving antenna in the same phase was assumed. In addition, a transmission correlation coefficient μ and a reception correlation coefficient ρ were introduced according to the Kronecker model as parameters representing the correlation in the transmission / reception antenna. The transmission correlation coefficient μ and the reception correlation coefficient ρ take values of 0 ≦ μ and ρ ≦ 1, and when 0, there is no correlation, and when 1, the same propagation path is obtained.

FIG. 6 shows the interference suppression amount _Th vs. BER characteristics when the modulation scheme is BPSK, E _b / N ₀ (Energy per bit to Noise power spectral density ratio) is 18 dB, and the transmission correlation coefficient μ is 0.3. It is shown.

The line-of-sight status of a propagation path in mobile communication always changes. In particular, when microcells are made to use higher frequency bands in the future, the line-of-sight situation may change frequently. Further, when the antenna interval is different for each receiving terminal, the reception correlation changes. Therefore, it is preferable to uniquely determining regardless interference suppression amount T _h of the magnitude of sight and out and correlations. Thus for example, the BER performance by most environmental setting a point to start pulling the floor as the interference suppression amount T _h.

Hereinafter, in the present embodiment is set to 10dB to _{T h} from FIG. From 6 this time, if you do not set the interference suppression amount _{T h} (i.e. _T h = 0 dB) in comparison with, BER than to set 10dB is understood that it is possible to improve one of the maximum 10 minutes.

FIG. 7 shows the received correlation coefficient ρ pair in the LOS environment (K = 3 dB) obtained by computer simulation when the modulation scheme is BPSK, E _b / N ₀ = 18 dB, and transmission correlation coefficient μ = 0.3. It is a BER characteristic. Similarly, FIG. 8 shows the reception correlation coefficient ρ vs. BER characteristics in the NLOS environment.

  From the figure, in the first method for suppressing the residual interference component after MMSE (marked with a circle in the figure), the BER characteristic deteriorates as the correlation increases, whereas the interference is suppressed without performing MMSE. It can be seen that the second method (Δ mark in the figure) always shows a constant characteristic regardless of the correlation, and shows a better characteristic than the first method in an environment where the correlation is large.

  In addition, in the method of the present invention (marked with □ in the figure), a method in which the SINR is good in units of subcarriers is selected. It can be seen that good BER characteristics can be obtained.

  The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and the design and the like within the scope of the present invention are also within the scope of the claims. include.

  The communication device described above may be any communication device such as a base station device, a terminal station device, a wireless device, a mobile terminal station device, and a mobile phone.

300 Transmitter 301a, 301b Encoding unit 302a, 302b Modulation unit 303 Pilot generation unit 304a, 304b Pilot multiplexing unit 305a, 305b Frequency time conversion unit 306a, 306b GI insertion unit 307a, 307b Radio transmission unit 308a, 308b Transmit antenna unit 400 Receiver 401a, 401b Reception antenna unit 402a, 402b Radio reception unit 403a, 403b GI removal unit 404a, 404b Time frequency conversion unit 405a, 405b Pilot separation unit 406 Propagation path estimation unit 407 MIMO separation unit 501 MMSE weight calculation unit 502 MMSE synthesis Unit 503 SINR calculation unit 504a, 504b method selection unit 505a, 505b demapping unit 408a, 408b decoding unit

Claims (7)

A communication device that receives a plurality of space-division multiplexed streams with a plurality of receiving antennas,
A propagation path estimator for estimating a propagation path matrix and noise power;
A weight calculator that calculates a weight matrix based on the propagation path matrix and noise power;
A synthesis unit that synthesizes a signal for each stream from the reception signal received by the reception antenna and the weight matrix;
A method selection unit for selecting one of the received signal and the synthesized signal;
A demapping unit that calculates likelihood information of the selected signal based on the propagation path matrix or the propagation path matrix and the weight matrix;
A decoding unit that performs error correction decoding based on the likelihood information;
A communication apparatus comprising:   The communication apparatus according to claim 1, wherein the weight calculation unit calculates an MMSE weight matrix for separating the received signal into the plurality of streams based on a minimum mean square error criterion.   The demapping unit calculates the likelihood information on the basis of the occurrence probability distribution of the residual interference component after combining for each stream when the signal for each stream combined by the method selection unit is selected. The communication apparatus according to claim 1.   When the received signal is selected by the method selection unit, the demapping unit is an interference component other than a stream for which likelihood information is calculated from a plurality of streams superimposed on the received signal. The communication apparatus according to claim 1, wherein the likelihood information is generated based on an occurrence probability distribution related to a stream of the first stream. Based on the weight matrix, the propagation path matrix, and noise power, a first signal-to-interference and noise power ratio is calculated for each combined stream, and the received signal is calculated based on the propagation path matrix and noise power. An SINR calculator that calculates a second signal-to-interference and noise power ratio for each of the streams,
When the first signal-to-interference and noise power ratio is larger than the second signal-to-interference and noise power ratio, the method selection unit selects the combined signal for each stream, and When the signal-to-interference and noise power ratio is larger than the first signal-to-interference and noise power ratio, the received signal is selected, and the first signal-to-interference and noise power ratio and the second signal-to-interference are selected. 2. The communication apparatus according to claim 1, wherein, when the noise power ratio is equal, one of the combined signal for each stream and the received signal is selected.   The communication apparatus according to claim 5, wherein the SINR calculation unit calculates the first and second signal-to-interference and noise power ratios using interference signal power suppressed by the error correction decoding. . In a communication system including a transmitter that transmits a plurality of streams by space-division multiplexing from a plurality of transmission antennas, and a receiver that receives signals transmitted from the transmitter by a plurality of reception antennas,
The receiver
A propagation path estimator for estimating a propagation path matrix and noise power;
A weight calculator that calculates a weight matrix based on the propagation path matrix and noise power;
A synthesis unit that synthesizes a signal for each stream from the reception signal received by the reception antenna and the weight matrix;
A method selection unit for selecting one of the received signal and the synthesized signal;
A demapping unit that calculates likelihood information of the selected signal based on the propagation path matrix or the propagation path matrix and the weight matrix;
A decoding unit that performs error correction decoding based on the likelihood information;
A communication system comprising:

Images