JP2003304178A - Multiinput multioutput turbo receiving method and receiver thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve an error rate characteristic when there is a big difference between received power of signals from respective transmitters. <P>SOLUTION: Received energy per information bit of a received signal from each transmitter is calculated (S2), and equalization and decoding processing is applied to the received signal from the receiver in the order of large received energy. That is, for example, only a received signal corresponding to a transmitter with large received energy in the previous prior information is used to perform equalization processing, and equalization and decoding processing is also applied to a received signal from a transmitter with small received energy according as the repetition of equalization and decoding is increased. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention receives signals from two or more integer N transmitters which use the same frequency band and transmit at the same time by one or more integer M antennas to perform equalization. The present invention relates to a multi-input multi-output (MIMO) turbo reception method and a receiver for separating signals from N transmitters by repeating decoding.

[0002]

2. Description of the Related Art The problem of the mobile communication business is how to accommodate a large number of users with high quality on a limited frequency. As a means for solving such a problem, there is a multi-input multi-output (MIMO) system. In this system, as shown in FIG. 1, a plurality of transmitters S1 to SN use the same frequency band and transmit at the same time, and these transmission signals are transmitted to a plurality of antennas A.
Received by the MIMO turbo receiver 10 including 1 to AM,
The MIMO turbo receiver 10 performs equalization decoding processing on the received signal, estimates the transmission symbols of the respective transmitters S1 to SN, and outputs them to the output terminals Out1 to OutN separately.

A multi-input multi-output turbo reception method is proposed in Japanese Patent Application No. 2001-043213 "Multi-input multi-output turbo reception method and channel estimation method". FIG. 3 shows the configuration of a transmitter that enables the multi-input multi-output turbo reception method proposed in this patent application. The channel encoder 11 performs error correction coding on the information bits, and an interleaver 12 interleaves (rearranges) the coded bit sequence (code error sequence) obtained by the error correction coding, and then the modulation means 13 performs a carrier wave. Modulate the signal and transmit. At this time, the training symbol sequence generated by the training symbol sequence generation means 14 is transmitted through the multiplexing means 15 and the encoded bit sequence is subsequently transmitted through the multiplexing means 15.

In the multi-input multi-output turbo receiver 10 shown in FIG. 1, the signals received by the antennas A1 to AM are not shown in the figure, but after reception amplification, they are converted into baseband signals, and further, for example, symbols (bits) are used. ) The signal is sampled at a cycle or an integral fraction thereof, and is input to the multi-input multi-output (MIMO) equalization unit 22 from the input terminals 21 _{1 to} 21 _M as a reception signal of a digital signal. The received signal from each antenna is also input to the channel estimation unit 23, and the same training symbol sequence as the training symbol sequence from each of the transmitters S1 to SN is input from the training symbol sequence generation unit 24 to this channel estimation unit 23. Each transmitter Sn
The transmission path characteristics between (n = 1, ..., N) and the receiving antennas A1 to AM, that is, impulse responses (channel values) are respectively estimated, and these estimated channel values correspond to the respective transmitter signals. Sequence decoding unit 25 _{1 to} 25
A second pre-information sequence lambda ₁ ^P to [lambda] _N ^P than _N is input to the MIMO equalization section 22, an equalization processing for all the input sequence is performed collectively, for each the respective transmitter signal (user) It is supplied to the signal sequence decoding units 25 _{1 to} 25 _N.

The structure of the MIMO equalizer 22 is shown in FIG.
The MIMO equalization unit 22 receives the received signal, and each decoding unit 25 ₁ to 2
Prior information sequence λ fed back from each 5 _N SISO decoder
₁ ^{P to} λ _N ^P , and the channel estimation value from the channel estimation unit 25 is input. If necessary, the offset compensating means 26 compensates for the difference between the transmitters and the antennas of the received signal, and the respective equalizing means 27 _{1 to} 27 _N for the respective sequences use the second prior information sequences λ ₁ ^{P to} λ _N ^P. And the channel estimation value, the interference replicas in the received signal in the sequence (transmitter signal) are regenerated by the interference replica generation unit 28, and subtracted from the received signal by the subtraction unit 29, so that intersymbol interference (ISI) , Reduce inter-channel interference (MAI). Furthermore, the ISI and MA left from the subtraction unit 29, respectively.
I is processed by the minimum mean square error (MMSE) filter 31, and the a priori information calculation unit 32 calculates the first a priori information sequence of each symbol with respect to the filter processing result and outputs the result. The prior information calculation unit 32 uses the MMSE filter 3
The log likelihood of the probability that the information bit (symbol) b (i) of the received signal for each transmitter is +1 and the probability that it is -1, with the output of 1, the channel estimation value, and the second prior information of the previous processing input The power ratio Λ ₁ is calculated, and the second prior information of the previous processing of the transmitter is subtracted from the Λ ₁ and output as the first prior information indicating the reliability of the equalization processing result.

As shown in FIG. 3, the first a priori information sequence for each user (transmitter) from the MIMO equalization unit 22 is reversely rearranged (de-interleaved) by the de-interleaver 33 in each sequence decoding unit 25 _{1 to} 25 _N. As an interleaved sequence, SISO (Si
ngl Input Singl Output) is input to the decoder 34 and SI
SO decoding (error correction decoding) is performed. The SISO decoder 34 outputs the decoding result of the error correction decoding process and the second prior information sequence indicating the reliability thereof. Each second
The prior information sequences are interleaved by the interleaver 35 and input to the MIMO equalization unit 22. The MIMO equalization process and the SISO decoding process are repeated a plurality of times for the same received signal, and the final decoding result is output. The control unit 36 controls and repeats each of these units.

The k-1th time of this iterative equalization decoding processing
The k-th process is shown in FIG. Here, user n (n = 1,
, N) equalization and SISO decoding unit 37n transmits
For the received signal from the device Sn, that is, for the signal sequence n.
For the sequence equalizer 27n and the signal sequence n in FIG.
3 is a combination of the sequence decoding units 25n in FIG.
Inputs and outputs of the above show only the second prior information, and other inputs and outputs
Power is omitted. Each obtained by the (k-1) th processing
From the equalization and SISO decoding unit 37n for user n
Second prior information λ_n ^PIs equalization and SI in the k-th processing
SO decryption unit 37 ₁~ 37_NIs supplied to all of and repeated
Processing is performed. In this way, from the first processing all
Equalization decoding processing for received signals from the transmitters S1 to SN
Was going on.

[0008]

In the MIMO turbo receiver described above, when there is a large difference between the received powers from the transmitters S1 to SN, the received signal of the transmitter (user) having a weak received power has a strong received power. More interference will be received from the received signal of the transmitter (user). As a result 1
The second prior information sequence generated after the second equalization decoding process has a large error, and if the equalization process is performed using this second prior information sequence, the bit error rate (BER)
The characteristics deteriorate.

[0009]

According to the present invention, the received energy per information bit of the received signal from each transmitter is estimated, and the first equalization (decoding) process is performed by at least one transmitter in descending order of received energy. The equalization (decoding) process is performed on the received signal from, and the equalization (decoding) process is performed on the received signal from the transmitter having the smallest received energy at least in the final equalization (decoding) process.

[0010]

DESCRIPTION OF THE PREFERRED EMBODIMENTS <First Embodiment> FIG. 6 shows a first embodiment of the present invention, in which parts corresponding to those in FIG. 3 are designated by the same reference numerals. In this embodiment, the second a priori information sequences output from the respective sequence decoding units 25 _{1 to} 25 _N are input to the MIMO equalization unit 22 via the selectors 41 _{1 to} 41 _N , respectively, and the transmitters S _{1 to S 1} are transmitted. The energy per information bit of the received signal from the SN is calculated by the received energy calculation unit 4
The selection control unit 43 controls the selectors 41 _{1 to} 41 _N based on the calculated received energy.
In addition, the MIMO equalization unit 22 and each sequence decoding unit 25 ₁ to
25 _N can have the same configuration as the conventional one.

FIG. 7 shows a process of iterative equalization and decoding processing from the (k-1) th time to the kth time according to the first embodiment.
It shows in correspondence with FIG. k-1st equalization and SISO
The second pre-information sequence obtained in the decoding unit 37 ₁ is supplied via the selector 41 ₁ to k-th all equalization and SISO decoding unit 37 ₁ to 37 _N, with equalization and SISO decoder 37 ₂ The obtained second prior information sequence is the selector 41 ₂
Through k equalization and SISO decoding section 37 ₁
~ 37 _N. Similarly, each equalization and SISO
The second a priori information sequence obtained by the decoding unit is supplied to each of the k-th equalization and SISO decoding units 37 ₁ through the selector.
~ 37 _N.

The selection control unit 43 performs selection control for the second prior information sequence based on the following standard, for example. If X (X = 1, 2, ..., N) is assigned to the users (transmitters) in order of increasing received energy per information bit, k
In the second iteration of equalization and error correction decoding, only the second a priori information series of the users in the order X satisfying k> X are selected, and the values of the second a priori information series of other users are all zero. Here, the received energy per information bit of each user is, for example, the antenna of the transmitter Sn and the antennas A1 to A1 of the receiver.
This can be obtained by integrating the square of each impulse response (channel estimation value) between AMs, taking the sum, and multiplying this by 1 / (coding rate).

The input terminals 21 ₁ to the antenna A1~AM as indicated in the received energy, for example, in FIG
A reception signal that has passed through the 21 _M, from the training symbol sequence generation unit 24 _M, respectively determined correlation by the correlation unit 45 ₁ to 45 _M of the training symbol sequence and the same training symbol sequence transmitter Sn to transmit these enter the correlation output to the shift register 46 ₁ -46 _M, respectively, by multiplying the weight multiplication unit to an output of the shift stages of the shift register 46 ₁ -46 _M, the addition unit 47 _1-4
7 _M respectively, that is, weighted addition is performed on the correlation output for a certain period of time, the received energy index is obtained, and the received energy index is further added by the adder 48, and the maximum value thereof is detected by the maximum value detector 49. This may be used as the received energy per symbol of the received signal from the user n (transmitter Sn).

An example of the processing procedure of the first embodiment shown in FIG. 6 will be described with reference to FIG. In step S1, the transmission channel characteristic between each transmitter and the receiver, that is, the channel value (impulse response) is estimated by the channel estimator 23, and in step S2 the received energy per information bit of the received signal from each transmitter is received. The energy is calculated by the energy calculator 42.
In step S3, the transmitters are sequentially numbered X (X = 1, 2, ..., N) in descending order of the calculated received energy. In step S4, the process number parameter k is initialized to 1. In step S5, the second prior information sequence of the transmitter of order X that satisfies k> X is selected. That is, the selector 41 _X corresponding to the prior information of the transmitter to be selected can be passed. Using the second prior information and the channel estimation value selected in step S6, equalization decoding processing is performed on the received signal from the transmitter in order X.

When k = 1, all the second prior information is not selected, so that the replica generating means 28 in FIG. 4 does not generate an interference replica signal. Therefore, all transmitters S1 to SN
The equalization decoding process is performed on each received signal from. 2nd of X = 1 transmitter starting to k = 2
Prior information is selected, an interference replica signal is generated using this information, this interference replica signal is subtracted from the received signal, and equalization decoding processing is performed on the rest, that is, from the transmitter S1 to the received signal, etc. The decryption process is performed. Similarly, as the number k of repeated processes increases, the number of second pieces of second prior information to be selected increases, but the increasing order is selected from the one having the largest received energy, so that the interference replica signal to be generated may be greatly erroneous. Without
Thus, the equalization decoding process can be performed correctly. Note that generally, the second prior information of all X transmitters satisfying k> X is selected, but if the number of processing times k is increased, the number of second prior information to be selected is not necessarily increased. The number of pieces of second prior information to be selected may be increased while selecting the number of pieces of second prior information, that is, every time the iterative process is performed twice. The equalization decoding process may be performed only on the received signals from the transmitters for which the second prior information is selected, but the process may be performed on the received signals from all the transmitters every time.

In step S7, k is the predetermined number of processing times k
_It is checked whether or not _R is obtained, and if it is not k _R , k is incremented by 1 in step S8 and the process returns to step S5. The value of k _R is set to be at least larger than N + 1, that is, when k = k _R , the equalization decoding process for the received signal from the transmitter having the smallest received energy is sufficiently repeated. If k = k _R in step S7, step S
At 9 then, the decoding result for the received signal from each transmitter obtained is output. As shown by the broken line in FIG. 9, as a result of the equalization decoding process in step S6 in step S10, the subsequent equalization decoding process for the received signal from the transmitter that was able to correctly decode all signals is stopped. You may.

As can be understood from the above description, the point is that the selective equalization decoding process is performed from the received signal from the transmitter having a large received energy. Not only when
For example, the control unit 36 controls so that only the sequence decoding units 25 _{1 to} 25 _N (FIG. 6) and the sequence equalization units 27 _{1 to} 27 _N (FIG. 4) that perform equalization decoding processing are operated. May be. In this case, the processing procedure is such that the "preliminary information" portion is deleted as shown by the broken line in FIG. 9, the selectors 41 _{1 to} 41 _{N in} FIG. 6 are omitted, and k > X in step S5. That is, that is, the equalization decoding process may be performed for the one with the largest received energy at k = 1.

<Second Embodiment> FIG. 10 shows a second embodiment of the present invention.
In the embodiment, parts corresponding to those in FIG.
Show. The difference from the first embodiment is that the sequence decoding unit 25₁~
25_NIs omitted, and instead, each transmission of the MIMO equalization unit 22 is omitted.
The equalized output of the received signal from the transmitter, that is, each MMSE filter
Output the decoding result by hard-deciding the output of the filter 31 (Fig. 4).
Hard decision section 51 ₁~ 51_NAnd the MIMO equalizer
The first prior information of each transmitter from 22 is the selector 41₁~ 4
1_NIs input to the MIMO equalizer 22 through
It The processing procedure of the receiver shown in FIG. 10 is shown in FIG.
Then, the first prior information is used as the prior information, and
Only the equalization process is performed instead of the equalization decoding process in step S6.
Then, in step S7, the number of processing iterations k is k_Rbecome
Then, in step S9, the MIMO equalizer 22 at that time
Hard decision of equalization output for received signals from each transmitter
Part 51₁~ 51_NHard decision is made by
Is output. Also in this case, the selector 41₁~ 41_NOmit
Then, the sequence equalization means 27 in the MIMO equalization unit 22₁~ 2
7_NK in>Equalization means 27 for series satisfying X_XOnly works
You may let me. In this second embodiment, error correction decoding is performed.
Therefore, the processing of step S10 cannot be performed.
Yes.

<Third Embodiment> In a third embodiment of the present invention, users (transmitters) are divided into a plurality of groups in the order of the magnitude of received energy, and selection processing is performed for each group. Here, for simplification, the order in which the energy per information bit is large is defined as users 1, 2, ..., N, that is, transmitters S1, S2 ,. As a method of classifying into groups, for example, groups G1 of users (transmitters) each having a number N / G obtained by dividing the number N of users by the number G of groups in descending order of energy per information bit.
, GG, or a group G1 of users received with energy per information bit equal to or more than (energy per information bit of user 1) × ((G−1) ÷ G), and in the following order (( The remaining users received with the energy per information bit of the user 1) × ((G−X) ÷ G) or more may be set as the group GX. Then, in the repetition of the equalization and decoding processing, in the kth iteration of the equalization and error correction decoding, only the signals of the users belonging to the group GX with k> X are selected, and the prior information of the users belonging to the other groups is selected. The values of series 2 are all 0.

The functional configuration of the turbo receiver in this case is shown in FIG.
11 is similar to that shown in FIG.
Shown in. In steps S1 and S2, channel estimation and reception energy estimation are respectively performed in the same manner as the procedure shown in FIG. In step S3, the received signals from the user (transmitter) are divided into G groups in descending order of received energy. As the method for this group, for example, the method described above can be adopted, and this is performed by the control unit 36. Further, in step S5, the prior information belonging to the transmitter group GX satisfying k> X is selected, and the equalization decoding process is performed on the received signal from the corresponding transmitter using the prior information selected in step S6. It is similar to the case of FIG. The processing in steps S7 and S8 is similar to that in FIG. 9, and the output in step S9 is also similar to S9 in FIG.

Also in this case, instead of selecting the second prior information,
For received signals from transmitters belonging to the selected group
Sequence decoding unit 25 so that only equalization decoding is performed.₁~
25 _NAnd series equalization means 27₁~ 27_NCorresponding in
You may make it operate only one thing. Second implementation
Similar to the embodiment, MIMO equalization without performing error correction decoding
Selector 41 for the first prior information from section 22₁~ 41_NSelected by
It is selected and returned to the MIMO equalizer 22 to perform repeated equalization.
In the same way, if you perform only the equalization process,
Equalizing means 27₁~ 27_NBy group selection
It may be operated. From the above description of the first to third embodiments
As will be appreciated, in this embodiment, the repetition
Received signal with high received energy at the beginning of processing
No. received energy is small after that
Including those, or excluding those that have been error correction decoded
Iterative processing was performed, and this method caused a large error.
It is possible to avoid the adverse effect of the interference replica signal.

<Fourth Embodiment> FIG. 12 shows a fourth embodiment of the present invention.
The structure of the embodiment is shown, and the portions corresponding to those in FIG. 6 are denoted by the same reference numerals. Selectors 41 _{1 to} 41 _N in FIG.
Instead of the weighting units 53 _{1 to} 53 _N , the selection control unit 43
A weight coefficient control unit 54 is provided instead of the. FIG. 13 shows the steps of the iterative equalization and decoding processing from the (k−1) th time to the kth time in the fourth embodiment, with the same reference numerals being given to the portions corresponding to those in FIG. 7. In the fourth embodiment, in the k-th equalization decoding process, the second prior information sequence obtained in the (k-1) th time is weighted by the weighting units 53 _{1 to} 53 _N. The weighting is controlled by the weighting coefficient control unit 54 so that the weaker the received energy, the smaller the weighting of the signal sequence with respect to the second prior information sequence. Therefore, this weight is a function of the received energy estimation value of all users and the number of repetitions. That is, as the number of repetitions increases, the difference between the weights given to the respective signal sequences becomes smaller, and finally the difference in weight becomes 0.
It is said that It is considered that the optimum weight depends on the equalization method, the SISO decoding method, and the error correction code, and it is necessary to obtain the optimum weight one by one by simulation or the like. It is advisable to provide a weight table 56 in the control unit 36 for reading out the received energy and the weight for the number of repetitions thus obtained.

Further, in the fourth embodiment, if the weighting coefficient is set to 0 or 1, and the same selection criteria as those used in the first to third embodiments are used, the same ones as in the first to third embodiments are obtained. As can be seen from the above, the first to third embodiments can be said to be a special case of the fourth embodiment. FIG. 14 shows an example of the processing procedure of the turbo receiver shown in FIG. Similar to FIG. 9, channel estimation and reception energy estimation are performed in steps S1 and S2. In this FIG. 12, k is initialized to 1 in step S3, and in step S4, the transmitters of the respective transmitters are associated with k and the reception energy. The weight for the prior information is determined by referring to the weight table 56 or using a predetermined function, and in step S5, the determined prior weight is multiplied by the prior information of the corresponding transmitter in the weight assigning unit 53n. And give weight.

In step S6, the received signal from each transmitter is subjected to equalization decoding processing using the channel estimation value and weighted prior information, and then k = k _R is checked to determine whether k = k _R.
If it is not _R , k is incremented by 1 and the process returns to step S4. If k = k _R , it is the same as outputting the decoding result in step S9. Also in this case, the error correction decoding is not performed, and a weight is given to the first a priori information, which is returned to the MIMO equalization unit 22 and only the equalization process is repeated. In this case, when k = k _R , the hard decision units 51 _{1 to} 51 _N make hard decisions on the equalized outputs of the MIMO equalization unit 22 and output them as decoding results.

In the case of performing error correction decoding in FIGS. 9, 11, and 14, instead of checking k = k _R in step S7, it is checked whether all received signals from all transmitters can be correctly decoded. If the decoding is successful, the iterative process may be stopped. The receiver 10 may have only one receiving antenna. In order to show the effect of the first embodiment, characteristic evaluation was performed by computer simulation. The parameters used in the simulation are as follows. Number of paths 5 Number of simultaneous transmission users 2 Number of reception antennas 3 Equalizer SC / MMSE decoding algorithm Max-Log-MAP constraint length 3 FIG. 15 shows a case where a desired user and one user who is another interference source perform transmission at the same time. Bit error rate of desired user (BE
R) shows the characteristics. Here, DUR indicates a desired wave to interference wave power ratio (db). From the figure, it can be seen that the BER characteristics deteriorate with the conventional method as the DUR decreases. This means that the small DUR means that the interference wave power (power of user 2) is larger than the desired wave power (power of user 1), and the desired user 1
This is because the received signal from the terminal receives more interference.
However, according to the first embodiment, it can be seen that the BER characteristic can be greatly improved in the negative DUR region. This means that the equalization decoding process is performed only on the signal of the user 2 and, as a result, even when the equalization decoding process of the signal of the user 1 is performed, the prior information sequence including many errors is not used, and the power consumption of the power is reduced. This is because the influence from large interference waves is reduced.

[0026]

As described above, the first to third aspects of the present invention are described.
According to the embodiment, the equalization (decoding) process is not performed on the received signal sequence having the small received energy at the beginning of the repetition, but the received signal sequence having the large received energy receives a large amount of interference, and the pre-processing including many errors is performed. Without using an information sequence, a priori information sequence with few errors is created for a received signal sequence with a large received energy, and the received energy with a small received energy can be created relatively accurately for a received signal sequence with a small received energy. Since the equalization (decoding) process is performed on the received signal sequence, the influence of interference can be reduced and the bit error rate characteristic can be improved. According to the fourth embodiment of the present invention, by giving a small weight to the prior information of the received signal sequence having small received energy, the adverse effect of the prior information sequence including many errors in the equalization process is reduced. Therefore, the error rate characteristic can be similarly improved.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration example of a multi-input multi-output simultaneous transmission / reception system.

FIG. 2 is a diagram showing a functional configuration example of a transmitter in FIG.

FIG. 3 is a diagram showing a functional configuration example of a conventional multi-input multi-output turbo receiver.

4 is a diagram showing a functional configuration example of a MIMO equalization unit 22 in FIG.

FIG. 5 is a diagram for explaining a conventional MIMO turbo equalization iterative process.

FIG. 6 is a diagram showing a functional configuration example of the first embodiment of the present invention.

FIG. 7 is a diagram for explaining repetitive processing of the first embodiment.

FIG. 8 is a diagram showing a specific example of a received energy calculation unit 42 in FIG.

FIG. 9 is a flowchart showing an example of a processing procedure of the first embodiment.

FIG. 10 is a diagram showing a functional configuration example of a second embodiment of the present invention.

FIG. 11 is a flowchart showing an example of the processing procedure of the third embodiment of the present invention.

FIG. 12 is a diagram showing a functional configuration example of a fourth embodiment of the present invention.

FIG. 13 is a diagram for explaining repetitive processing of the fourth embodiment.

FIG. 14 is a flowchart showing an example of the processing procedure of the fourth embodiment.

FIG. 15 is a diagram showing an electronic computer simulation result for explaining the effect of the first embodiment.

Continued front page    (72) Inventor Takahiro Asai             2-11-1, Nagatacho, Chiyoda-ku, Tokyo Stock             Ceremony company NTT Docomo (72) Inventor Shigeru Tomisato             2-11-1, Nagatacho, Chiyoda-ku, Tokyo Stock             Ceremony company NTT Docomo F-term (reference) 5K022 FF00                 5K046 AA05 EE47 EF03 EF13

Claims (12)

[Claims] 1. A transmitter that receives signals from an integer N of two or more transmitters and repeats equalization processing for the received signals by using a channel estimation value between the transmitters and the receivers. In the receiving method that separates and decodes the signal from, the reception energy per information bit of the reception signal from each transmitter is obtained, and the reception signal from the transmitter with a large reception energy is selected at the beginning of the repetition of the equalization process. A multi-input multi-output turbo reception method characterized by performing equalization processing, and at the end of repetition of the equalization processing, selective equalization decoding processing is performed on a received signal from a transmitter with small received energy. 2. The selective equalization process is performed on a received signal from a transmitter having a smaller received energy as the equalization process is repeated.
The described multi-input multi-output turbo reception method. 3. The X-order is the order in which the received energy is large, and the k-th equalization process is a selective equalization process for the received signal from the X-th transmitter that satisfies k> X. The described multi-input multi-output turbo reception method. 4. The descending order of received energy,
N _G1 , N _G2 , N _G3 , ..., N _GX pieces (N _G1 , N _G2 , ..., N _GX
Is a natural number) and the received signals from the N transmitters are classified into groups G1, G2, ..., GX, and the k-th equalization process selects the received signals from the transmitters in the group GX that satisfy k> X. The multi-input multi-output turbo reception method according to claim 1, wherein equalization processing is performed. 5. For each equalization process, the received signal from each transmitter subjected to the equalization process is subjected to error correction decoding, and prior information indicating the reliability thereof is obtained, and the above equalization process is performed using the above prior information. The equalization decoding process is repeated, and the selective equalization process is performed by selecting only the prior information obtained by the previous equalization decoding process of the reception signal from the transmitter which is the target of the selective equalization process. The multi-input multi-output turbo reception method according to any one of claims 1 to 4, wherein the multi-input multi-output turbo reception method is used in a digitization process. 6. For each equalization process, prior information indicating the reliability of the equalization output of the received signal from each transmitter is obtained, the equalization process is performed including the prior information, and the selection is performed. 5. The equalization process uses only the prior information obtained by the previous equalization process of the reception signal from the transmitter which is the target of the selective equalization process, in the selective equalization process. The multi-input multi-output turbo reception method according to any one of 1. 7. A signal from an integer N number of transmitters equal to or greater than 2 is received, and a channel estimation value between the transmitter and the receiver and prior information obtained in the previous processing are received for the received signal. Equalization processing is performed using the equalization processing, error correction decoding is performed on the result of the equalization processing, prior information indicating the reliability thereof is obtained, and the above equalization decoding processing is repeated to separate and decode the signals from each transmitter. In the method, the received energy per information bit of the received signal from each transmitter is obtained, the larger the received energy is, the greater the weight is given to the prior information used in the equalization process, and the repetition of the equalization decoding is performed. A multi-input multi-output turbo reception method, characterized in that the difference in weight given to the prior information is reduced as the process proceeds. 8. A signal from an integer N number of transmitters equal to or greater than 2 is received, and a channel estimation value between the transmitter and the receiver and prior information obtained in the previous process are received for the received signal. A receiving method in which the equalization output of each transmitter is equalized by using the equalized output and the prior information indicating the reliability thereof is output repeatedly, and the signal from each transmitter is separated and decoded from the equalized output. In, the received energy per information bit of the received signal from each transmitter is obtained, and the greater the received energy, the greater the weight given to the prior information used in the equalization process, and the repetition of the equalization process proceeds. The multi-input multi-output turbo reception method is characterized in that the difference in weight given to the prior information is reduced according to the above. 9. A receiver for receiving signals from an integer N transmitters of 2 or more, wherein the received signal and a training signal corresponding to each transmitter are input, and between the transmitters and the receivers. A channel estimation unit for estimating each channel value of, a received signal, each channel value, and prior information for each transmitter are input, and a received signal from each transmitter is equalized and output. An input multi-output equalization unit, a decoding unit that receives the outputs of the multi-input multi-output equalization unit, performs error correction decoding, and outputs a priori information as information indicating the reliability thereof, and each of the decoding units Preliminary information for each transmitter is entered,
A selection unit that inputs only selected ones to the multi-input multi-output equalization unit, a reception energy calculation unit that calculates reception energy per information bit of the reception signal from each transmitter, and for the same reception signal While repeatedly operating the multi-input multi-output equalization unit and the decoding unit, as the repetition is repeated, the selection unit is selected so as to sequentially select the prior information of the transmitter with the small reception energy from the transmitter with the large reception energy. A multi-input multi-output turbo receiver including a control unit for controlling. 10. A receiver for receiving signals from an integer N number of transmitters equal to or greater than 2, wherein a received signal and a training signal corresponding to each of the transmitters are input, and between the transmitters and the receivers. A channel estimation unit for estimating each channel value of, the received signal, each channel value, and prior information for each transmitter are input, and the received signal from each transmitter is equalized and output. A multi-input multi-output equalization unit that outputs a priori information indicating the reliability, a decoding unit that inputs the output of the multi-input multi-output equalization unit and outputs a hard decision value, the multi-input multi-output, etc. Preselection information for each transmitter is input from the equalizer, and only a selected one is input to the multi-input multi-output equalizer, and a reception signal per information bit from the transmitter is received. With a received energy calculator that calculates energy The selection unit is configured to repeatedly operate the multi-input multi-output equalization unit with respect to the same reception signal, and to sequentially select the prior information of the transmitters with large reception energy from the transmitter with large reception energy as the repetition is repeated. A multi-input multi-output turbo receiver including a control unit for controlling the. 11. A receiver for receiving signals from an integer N transmitters of 2 or more, wherein the received signal and a training signal corresponding to each transmitter are input, and between the transmitters and the receivers. A channel estimation unit for estimating each channel value of, a received signal, each channel value, and prior information for each transmitter are input, and a received signal from each transmitter is equalized and output. An input multi-output equalization unit, a decoding unit that receives the outputs of the multi-input multi-output equalization unit, performs error correction decoding, and outputs a priori information as information indicating the reliability thereof, and each of the decoding units Preliminary information for each transmitter is entered,
A weighting unit that is weighted and is input to the multi-input multi-output equalization unit as the pre-information, a reception energy calculation unit that calculates reception energy per information bit of a reception signal from each transmitter, and The larger the received energy, the larger the weight is given to the prior information of the transmitter, the more weight is supplied to the weighting unit, and the multi-input multi-output equalization unit and the decoding unit are repeatedly operated for the same reception signal. , A multi-input multi-output turbo receiver including a control unit that reduces the mutual difference of weights given to prior information as the repetition is repeated. 12. A receiver for receiving signals from an integer N transmitters of 2 or more, wherein a received signal and a training signal corresponding to each transmitter are input, and between the transmitters and the receivers. A channel estimation unit for estimating each channel value of, the received signal, each channel value, and prior information for each transmitter are input, and the received signal from each transmitter is equalized and output. A multi-input multi-output equalization unit that outputs a priori information indicating the reliability, a decoding unit that inputs the output of the multi-input multi-output equalization unit and outputs a hard decision value, the multi-input multi-output, etc. Preliminary information for each transmitter from the equalization unit is input, a weight is assigned to the multi-input multi-output equalization unit and is input as the prior information to the multi-input multi-output equalization unit, and information on the received signal from each transmitter. Receive to calculate received energy per bit The energy calculation unit and a weight are supplied to the weight giving unit so that a larger weight is given to the prior information of the transmitter as the received energy is larger, and the multi-input multi-output equalization unit and the decoding unit are provided for the same received signal. A multi-input multi-output turbo receiver comprising: a control unit that repeatedly operates the units and reduces the mutual difference of weights given to the prior information as the repetition is repeated.