JP2011250073A - Communication system, transmitter, receiver, and communication method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a transmission system for reducing Modulo Loss.SOLUTION: A transmitter comprises: a coding part 3 generating code bits by error correction coding; a modulation part 17 generating a modulation signal by modulating transmission bits; a perturbation vector addition part 23 generating a transmission signal by adding a perturbation vector to the modulation signal; a perturbation bit calculation part 25 calculating perturbation bits based on the perturbation vector; and a perturbation bit addition part 7 calculating the transmission bits by adding the perturbation bits to at least a part of the code bits.

Description

    The present invention relates to mobile communication technology.

1. About THP
Tomlinson Harashima Precoding (THP) is a technique in which a transmission apparatus transmits a signal to a reception apparatus in a situation where interference exists, and it is assumed that the transmission apparatus has grasped the interference in advance. By performing both modulo (modulo) operations, both the transmission and reception devices can realize transmission and reception of signals in which an increase in transmission power is suppressed (see Non-Patent Document 1 below). A Modulo operation performed by the transmission / reception device in communication using THP will be described. The Modulo operation performed by the transmission device is a process that reduces the transmission power by keeping the transmission signal below a certain amplitude, and the Modulo operation performed by the reception device is a process that restores the signal that has been subjected to the Modulo operation on the transmission device side. is there. More specifically, the Modulo operation is an integral multiple of a known value τ on both sides of transmission and reception for the real part (In-phase channel, I-ch) and imaginary part (Qudrature channel, Q-ch) of the transmission signal, etc. This is an operation in which the signal falls within the range [−τ / 2, τ / 2]. An example of this Modulo operation is shown in FIG. In FIG. 23, the Modulo operation is represented by a process in which the signal represented by ● moves to the position of ○, and moves to ○ by adding (−2) τ + j * (− 1) τ to ●. (See arrow). Here, j is an imaginary unit. ○ is within the range of [-τ / 2, τ / 2] from the origin (0) for both real and imaginary parts. Thus, the Modulo operation has the effect of keeping the signal amplitude within a certain range. Usually, when the average power of the modulation symbol is normalized to 1, the Modulo width τ is a predetermined value known in advance on the transmission / reception side according to the modulation scheme. For example, τ = 4 for BPSK, τ = 2√2 for QPSK, τ = 8 / √10 for 16QAM, and τ = 16 / √42 for 64QAM.

となる。ここで、Re(x)はxの実部を表し、Im(x)はxの虚部を表す。floor(x)はx を超えない最大の整数を表す。 By this Modulo calculation, even in an environment where the receiving side receives a large amount of interference, a signal can be transmitted while suppressing an increase in transmission power due to interference removal. In addition, the Modulo operation is expressed as an equation:

It becomes. Here, Re (x) represents the real part of x, and Im (x) represents the imaginary part of x. floor (x) represents the largest integer that does not exceed x.

  Next, the principle of THP will be described. Let s be the desired signal and f be the interference. The transmitting apparatus first subtracts the interference f from the desired signal s. However, since the subtracted signal s −f usually has a large amplitude, if it is transmitted as it is, the transmission power increases. Therefore, the transmission device performs a Modulo operation on this signal s − f and transmits a signal indicated by Modτ (s − f).

となり、受信装置が所望信号を検出できる。このように受信側でもModulo演算を行うことにより、所望信号sを受信側で復元することができる。以上が、THPの仕組みである。 Then, the transmitter can keep the transmission signal in the range of [-τ / 2, τ / 2] from the origin for both I-ch and Q-ch, and it can save power compared to sending the signal s − f. A suppressed signal can be transmitted. Here, if the propagation path characteristic is set to 1 and the influence of noise is ignored, the received signal is Modτ (s−f) + f. This is because the receiving apparatus receives interference f. When the receiving device performs a Modulo operation on this received signal,

Thus, the receiving apparatus can detect the desired signal. As described above, the Modulo operation is also performed on the reception side, whereby the desired signal s can be restored on the reception side. The above is the THP mechanism.

2. MU-MIMO THP
As shown in FIG. 24, when a base station (BS) transmits a signal to a plurality of terminals (Mobile Terminal: MT) at the same time and the same frequency, usually, multi-user interference (Multi User Interference : MUI) occurs. Downlink (DL) MU-MIMO (Multi-User Multi Input Multi Output) is a method of multiplexing a plurality of MTs by removing the MUI efficiently using THP.

  DL MU-MIMO THP is a technology based on the premise that the BS knows all MT channel state information (Channel State Information: CSI). This is because, as described above, the THP requires that the BS that is the transmission device knows the interference that the MT that is the reception device receives, and the DL MU-MIMO THP uses CSI to calculate the interference. This is because it is necessary to use.

  In DL MU-MIMO THP, a method of precoding using QR decomposition based on the ZF norm and MMSE norm, and a method of precoding using BLAST based on the ZF norm MMSE norm are known ( Non-Patent Document 2 and Non-Patent Document 3 below).

3. MU-MIMO Vector Perturbation
There is also a technology called MU-MIMO Vector Perturbation (VP) as a technology similar to MU-MIMO THP (see Non-Patent Document 4). When performing MU-MIMIO VP, the BS generates a transmission signal by the following two procedures.

1. Add an appropriately multiple signal (perturbation vector) of the Modulo width to the desired signal.
2. Remove interference between MTs by the same process as linear precoding.

  At this time, the signal to be added is called a perturbation vector. If this perturbation vector is appropriately selected in consideration of signals addressed to all MTs and propagation path conditions, transmission power can be suppressed (Non-patent Reference 4). Although MU-MIMO VP uses the same linear filter as linear precoding, it performs non-linear processing such as addition of perturbation vectors, and is classified as non-linear processing similar to MU-MIMO THP. In addition, each MT can extract a desired signal by performing a Modulo operation similar to MU-MIMO THP on the received signal.

  In this MU-MIMO VP, as with the MU-MIMO THP, it is possible to more efficiently use the spatial resources of the multiple transmission antennas of the BS, and thus to improve the frequency utilization efficiency.

H. Harashima and H. Miyakawa, “Matched-Transmission Technique for Channels With Intersymbol Interference”, IEEE Transactions On Communications, Vol. Com-20, No. 4, pp. 774-780, August 1972. J. Liu and A. Krzymien, “Improved Tomlinson-Harashima Precoding for the Dowinlink of Multiple Antenna Multi-User Systems”, Proc. IEEE Wireless and Communications and Networking Conference, pp. 466-472, March 2005. M. Joham, J. Brehmer, and W. Utschick, “MMSE Approaches to Multiuser Spatio-Temporal Tomlinson-Harashima Precoding,” Proc. Of 5. ITG Conference on Source and Channel Coding, Erlangen,, pp. 387-394, January 2004. BM Hochwald, CB Peel, and AL Sindlehurst, “A vector-perturbation technique for near-capacity multi-antenna multi-user communication-part II: perturbation”, IEEE Trans. On Communications, vol. 53, pp. 537-544, March 2005. J. G. Proakis and M Salehi, “Digital Communications”, McGraw Hill Higher Education, 5th revised edition, January 2008.

  Even when the interference power is large, the THP can transmit a signal from which interference has been removed in advance while suppressing an increase in transmission power.

  However, at the time of reception, the points that shifted periodically in the real axis (I-ch) and imaginary axis (Q-ch) directions with the Modulo width and interval are all regarded as the same point, so the number of reception candidate points increased. The signal detection performance is degraded. This drop in performance is called Modulo Loss. For example, when QPSK is used as the modulation method, the result is as shown in FIG. The four marks (•••••• □) in FIG. 25 correspond to different bits, and the receiving side determines which of the 4 signal points has been transmitted. Now, it is assumed that the signal is shifted in the direction of the arrow shown in FIGS. 25A and 25B due to the influence of noise. In FIG. 25 (a), the point closest to the reception signal point indicated by x at the tip of the arrow is ◯, whereas in FIG. As described above, the signal detection performance deteriorates as the number of reception candidate points increases. This deterioration occurs not only in QPSK but in any modulation scheme.

  An object of this invention is to provide the transmission system which reduces Modulo Loss.

  According to one aspect of the present invention, a transmission device in a communication system in which a transmission device transmits a signal to at least one reception device, a coding unit that generates a code bit by error correction coding, and a transmission bit A modulation unit that generates a modulation signal by modulating, and a perturbation vector addition unit that generates a transmission signal by adding a perturbation vector that is a signal that is an integral multiple of a known magnitude signal in the reception device and the transmission device to the modulation signal And a perturbation bit calculation unit that calculates a perturbation bit based on the perturbation vector, and a perturbation bit addition unit that calculates a transmission bit by adding the perturbation bit to at least a part of the sign bit. A featured transmitter is provided. It is preferable that the perturbation bit calculation unit calculates a perturbation bit based on a value obtained by dividing the perturbation vector by the signal of the known magnitude. The perturbation bit calculator preferably calculates the perturbation bit based on whether the divided value is an odd number or an even number. It is preferable that an interference subtracting unit that subtracts an interference signal received by the reception signal from the modulation signal. Preferably, the apparatus further includes a coefficient multiplier that multiplies the interference signal received by the reception signal by a positive coefficient less than 1 from the modulated signal, and an interference subtractor that subtracts the signal multiplied by the coefficient. . Here, it is preferable that the perturbation vector addition unit generates the transmission signal by Modulo calculation.

  Further, the transmission device transmits a transmission signal addressed to a plurality of reception devices simultaneously at the same frequency, and a linear filter multiplication unit that multiplies the transmission signal calculated for the plurality of reception devices by a linear filter. You may make it have. Furthermore, a perturbation vector calculation unit that calculates the perturbation vector to be added to the receiving device based on the linear filter and the modulated signal transmitted to the receiving device may be provided.

  The present invention is also a receiving apparatus in a communication system in which a transmitting apparatus transmits a signal to at least one receiving apparatus, and estimates a perturbation bit based on the received signal to calculate a perturbation bit soft estimate. Based on a perturbation bit soft estimation unit, a transmission bit soft estimation unit that estimates a transmission bit based on a received signal and calculates a transmission bit soft estimation value, and the perturbation bit soft estimation value and the transmission bit soft estimation value And a decoding unit that performs decoding based on the code bit soft estimated value and a Check Node that calculates the code bit soft estimated value. The decoding unit further calculates a code bit soft estimate, and the CheckNode further calculates a transmission bit soft estimate based on the code bit soft estimate and the perturbation bit soft estimate, and the code bit It is preferable to newly calculate the perturbation bit soft estimation value based on the soft estimation value, the perturbation bit soft estimation value, and the transmission bit soft estimation value. In addition, the transmission bit soft estimation unit is the same reception signal point in that the reception signal is translated in the real part or imaginary part direction by an integral multiple of a signal having a known magnitude in the reception device and the transmission device. The perturbation bit soft estimator calculates the transmitted bit soft estimate by regarding the received signal as the same received signal point by translating the received signal in the real part or imaginary part direction by an even multiple of the known size. It is preferable to calculate the perturbation bit soft estimate by assuming that there is. The decoding unit preferably includes a coefficient multiplier that multiplies the received signal by a positive coefficient less than 1.

  In addition, the present invention is a communication system in which a transmission device transmits a signal to at least one reception device, a coding unit that generates code bits by error correction coding, and a modulation signal by modulating transmission bits. A modulation unit to generate, a perturbation vector addition unit that generates a transmission signal by adding a perturbation vector that is a signal that is an integer multiple of a signal of a known magnitude in the reception device and the transmission device, and the perturbation vector A perturbation bit calculation unit that calculates a perturbation bit based on the above, a perturbation bit addition unit that calculates a transmission bit by adding the perturbation bit to at least a part of the sign bit, and a received signal A perturbation bit soft estimator that estimates the perturbation bit and calculates the perturbation bit soft estimate based on the received signal, and the transmission bit soft estimation based on the received signal A transmission bit soft estimator for calculating the bit, a check node for calculating a code bit soft estimate based on the perturbation bit soft estimate and the transmission bit soft estimate, and decoding based on the code bit soft estimate A communication system comprising: a decoding unit; and a receiving device including the decoding unit.

  According to another aspect of the present invention, there is provided a transmission method in a communication system in which a transmission device transmits a signal to at least one reception device, the step of generating a code bit by error correction coding, A step of generating a modulation signal by modulation, a step of adding a perturbation vector that is an integer multiple of a predetermined magnitude to the modulation signal to generate a transmission signal, and a perturbation bit based on the perturbation vector And a step of calculating a transmission bit by adding the perturbation bit to at least a part of the code bit.

  Further, a reception method in a communication system in which a transmission device transmits a signal to at least one reception device, the step of estimating a perturbation bit based on the reception signal and calculating a perturbation bit soft estimation value; Estimating a transmission bit based on a signal and calculating a transmission bit soft estimate; calculating a code bit soft estimate based on the perturbation bit soft estimate and the transmission bit soft estimate; Decoding based on the code bit soft estimate.

  The present invention may be a program for causing a computer to execute the above transmission method or reception method, or a computer-readable recording medium for recording the program. The program may be acquired by a transmission medium such as the Internet.

  Furthermore, the present invention is a processor used in a transmission apparatus in a communication system in which a transmission apparatus transmits a signal to at least one reception apparatus, and a coding unit that generates code bits by error correction coding, and a transmission A modulation unit that modulates bits to generate a modulated signal, and a perturbation vector that adds a perturbation vector that is an integer multiple of a signal of a known magnitude in the receiving device and the transmitting device to the modulated signal to generate a transmission signal A vector addition unit, a perturbation bit calculation unit that calculates a perturbation bit based on the perturbation vector, and a perturbation bit addition unit that calculates a transmission bit by adding the perturbation bit to at least a part of the sign bit. It is a processor provided.

  Further, a processor used in a receiving apparatus in a communication system in which a transmitting apparatus transmits a signal to at least one receiving apparatus, and calculates a perturbation bit soft estimate by estimating a perturbation bit based on the received signal. A perturbation bit soft estimator, a transmission bit soft estimator that estimates a transmission bit based on a received signal and calculates a transmission bit soft estimation value, and is based on the perturbation bit soft estimation value and the transmission bit soft estimation value. The processor may include a Check Node that calculates a code bit soft estimate and a decoding unit that performs decoding based on the code bit soft estimate.

  ADVANTAGE OF THE INVENTION According to this invention, the transmission system which reduces Modulo Loss can be provided.

It is a functional block diagram which shows the example of 1 structure of the transmitter and receiving device by the 1st Embodiment of this invention. It is a flowchart figure which shows the flow of a process of a transmitter. It is a flowchart figure which shows the flow of a process of a receiver. It is the figure which showed the method by which a perturbation bit addition part calculates a transmission bit from a code bit and a perturbation bit by numerical expression as a generator matrix. It is a figure which shows the outline | summary of the calculation method of the transmission bit LLR in a transmission bit soft estimation part. It is a figure which shows the outline | summary of the calculation method of the transmission bit LLR in a transmission bit soft estimation part. It is a figure which shows one structural example of the base station apparatus in the modification 1. FIG. It is a figure which shows the example of 1 structure of the terminal (1st terminal) which receives the signal which is not DPC-encoded. It is a figure which shows one structural example of the terminal (2nd terminal) which receives the signal by which DPC encoding was carried out. It is a figure which shows one structural example of the base station apparatus in the modification 2. It is a figure which shows the structural example of the DPC encoding part and DPC decoding part which are the element parts of the 2nd Embodiment of this invention. It is a figure which shows the example of 1 structure of the DPC encoding part and DPC decoding part by the 3rd Embodiment of this invention. It is a figure which shows the example of the production | generation matrix which combined two transmission bits and perturbation bits in QPSK by which one transmission bit is arrange | positioned at a real part and an imaginary part corresponding to a real part and an imaginary part. It is a figure which shows a mode that the distance shown in FIG. 6 was generalized to the complex number plane. It is a flowchart figure which shows the flow of operation | movement of a DPC encoding part. It is a functional block diagram which shows the example of 1 structure of the communication apparatus by the 4th Embodiment of this invention. It is a figure which shows the example of 1 structure of the base station apparatus by this Embodiment. It is a figure which shows the example of 1 structure of the terminal station apparatus by this Embodiment. It is a functional block diagram which shows the example of 1 structure of the base station apparatus by the 5th Embodiment of this invention. It is a figure which shows the structure of the principal part of FIG. 19 in detail. In this Embodiment, it is a figure which shows the matrix corresponding to FIG. It is a figure which shows an example of set positioning corresponding to each modulation system. It is the figure which showed an example of Modulo calculation. It is a figure which shows a mode that a base station (Base Station: BS) transmits a signal with respect to several terminals (Mobile Terminal: MT) at the same time and the same frequency. It is a figure which shows the mode of the performance fall (Modulo Loss) at the time of using QPSK for a modulation system. It is a figure which shows the difference with the constellation with which a bit has a correlation, and the constellation without a correlation.

  The present invention provides a method for reducing Modulo Loss by using a gain of error correction coding by repetitive signal detection processing in a receiving apparatus in a communication system that performs THP and VP. This utilizes the so-called “turbo principle” used in turbo coding, turbo equalization, and the like. Further, in order for the receiving device to make the turbo principle function, the transmitting device newly performs certain processing on the transmission signal. Hereinafter, a communication technique according to an embodiment of the present invention will be described with reference to the drawings.

The “linear filter” in this specification may be, for example, Q after QR decomposition of H ^H (MU-MIMO THP) in the first to fourth embodiments, or the fifth embodiment. Then, for example, an inverse matrix of a propagation path matrix or the like may be used (MU-MIMO VP).

<First Embodiment>
In the first embodiment of the present invention, first, the operation of the element part of the present invention will be described using an AWGN (Additive White Gaussian Noise) channel. Secondly, a modification in which the element part is applied to DL MU-MIMO THP when there are two terminal devices will be described. Third, a modified example applied to a case where there are two or more arbitrary terminal devices will be described.

1. Element Part First, the operation of the element part of the present embodiment will be described. FIG. 1 is a functional block diagram showing a configuration example of a transmission apparatus according to the first embodiment of the present invention. As shown in FIG. 1, the transmitting apparatus A side of the element part is called a DPC encoding unit 1, and includes an encoding unit 3, an interleaving unit 5, a generator matrix storage unit 11, a perturbation bit addition unit 7, a modulation unit 17, an interference It comprises a subtractor 19, a dither sequence adder 21, a perturbation vector adder 23, and a perturbation bit calculator 25. The components from the modulation unit 17 to the perturbation vector addition unit 23 are referred to as a transmission signal generation unit 15.

  The configuration and operation of the transmission apparatus will be described below with reference to the functional block diagram of FIG. 1 and the flowchart of the transmission apparatus of FIG.

  (Step S <b> 101 in FIG. 2) The encoder 3 performs error correction encoding on the information bits, generates code bits, and inputs them to the interleaver 5. Note that the error correction code scheme may be, for example, a convolutional code, a turbo code, a low density parity check (LDPC) code, a repeat-accumulate (RA) code, or other coding scheme.

  (Step S <b> 102) The interleave unit 5 performs interleaving on the code bits, and inputs the interleaved code bits to the perturbation bit adder 7.

  (Step S103) Next, after the generation matrix storage unit 11 inputs the generation matrix to the perturbation bit addition unit 7, the perturbation bit addition unit 7 generates a perturbation bit for the sign bit input from the interleaving unit 5. The transmission bits are calculated by adding according to the rules represented by the matrix. Here, the perturbation bit is a bit calculated from the sign bit before the sign bit to which the perturbation bit is to be added, and is calculated by the transmission signal generation unit 15 and the perturbation bit calculation unit 25. A bit obtained by adding at least a part of the perturbation bits to the sign bit in the perturbation bit addition unit 7 is referred to as a transmission bit. The perturbation bit adding unit 7 inputs transmission bits to the modulation unit 17 in the transmission signal generating unit 15.

  A method in which the perturbation bit adding unit 7 calculates a transmission bit from the sign bit and the perturbation bit is shown as a generation matrix in FIG. The equation shown in FIG. 4 shows the relationship between the code bit, the transmission bit, and the perturbation bit when the encoded block length after encoding (the number of bits output by one encoding portion) is 9.

  As shown in the above equation (1-1), each transmission bit btxk (k = 1 to 9) is the same k-th code bit bcik and perturbation from 1 to k-1 in order before k. A bit selected from bits bpv1 to bpv (k-1) is added. The reason why only the perturbation bits in the previous order can be used is that the transmission signal generation unit 15 and the perturbation bit calculation unit 25 described later do not calculate the perturbation bits btxk to btx9 when the transmission bit btxk is calculated.

  Which perturbation bit (bpv1 to bpvk) to select and add when calculating each transmission bit btxk can be selected at random if it is shared in advance on the sending and receiving sides. What is a generator matrix. In the generator matrix shown in FIG. 4, a component containing 1 in the lower triangular component (portion indicated by P_low) of the 9 × 9 matrix portion in the right half indicates a perturbation bit to be added. Since the generator matrix is a lower triangular matrix with a diagonal component of 0, each transmission bit is always calculated using the perturbation bits in the order k before the sign bit.

  The perturbation bits to be added may be randomly selected for each transmission bit btxk. However, when the perturbation bits added between the transmission bits are compared, it is desirable that they be arranged as different as possible. This is because, as in other error correction coding schemes, by giving constraints with low correlation to each other as much as possible, a high gain can be obtained and the error rate can be reduced.

  For the sake of simplicity, FIG. 4 illustrates the case where there are nine code bits, transmission bits, and perturbation bits. However, the number is not limited to nine and may be larger. In particular, when the encoding block length in the encoding unit 3 is longer, it is desirable that the value is equal to the encoding block length. In the reception process, the decoding of the error correction code and the process corresponding to this step are repeated. When the unit for detecting the bits is prepared at that time, the information bits are detected for each coding block. Because it can be done. However, it does not necessarily have to be equal to the coding block.

  (Step S <b> 104) The modulation unit 17 generates a modulation signal from the transmission bits input from the perturbation bit addition unit 7 and inputs the modulation signal to the interference subtraction unit 19. In the present embodiment, it is assumed that the modulation signal is BPSK (Binary Phase Shift Keying).

  (Step S <b> 105) The interference subtraction unit 19 subtracts the interference signal from the modulation signal and inputs the modulation signal obtained by subtracting the interference signal to the dither sequence addition unit 21. A method for calculating interference will be described later.

  (Step S106) The dither sequence addition unit 21 adds a dither sequence, which is a known signal in the transmission / reception device, to the signal input from the interference subtraction unit 19, and inputs the added signal to the perturbation vector addition unit 23. . Here, the signal input to the perturbation vector adder 23 is called a perturbed signal. The dither sequence is a random signal for eliminating the bias of the signal after adding the perturbation vector. The transmission signal generation unit 15 preferably includes the dither sequence addition unit 21, but the dither sequence may not be added.

  (Step S107) The perturbation vector is a signal having, in its real part, a component that is an integral multiple of the Modulo width τ used in THP. In order to reduce the transmission power, a perturbation vector is appropriately selected so that the transmission signal falls within [−τ / 2, τ / 2], and the perturbation vector is added to the perturbed signal. Further, it is assumed that the Modulo width τ used in this THP is known not only by the transmitting apparatus but also by the receiving apparatus. In the present embodiment, the operation of the perturbation vector adder is expressed by the following equation because the transmission signal exists only in the real part because the modulation method is BPSK.

Here, floor (x) represents the largest integer not exceeding x. The perturbation vector addition unit 23 inputs the perturbation vector (z _k = −floor ((x + τ / 2) / τ) τ) to the perturbation bit calculation unit 25.

(Step S <b> 108) The perturbation bit calculation unit 25 converts the perturbation vector input from the perturbation vector addition unit 23 into a perturbation bit and inputs the perturbation bit to the perturbation bit addition unit 7. Here, the perturbation bit calculation unit 25 converts the perturbation bit to 0 when the perturbation vector z _k expressed by an integral multiple of the Modulo width is an even multiple of the Modulo width, and converts it to the perturbation bit 1 when the perturbation vector is an odd multiple.

  (Step S109) Finally, a signal after addition of the perturbation vector (after Modulo calculation) is transmitted to the receiving apparatus. However, here, a description will be given on the assumption that a signal (signal after DPC code) generated by adding the perturbation vectors in the perturbation vector addition unit 23 is transmitted to the receiving device. However, as will be described later, this is applied to MU-MIMO THP. In this case, a signal is transmitted to the receiving apparatus after performing transmission processing such as multiplication of a linear filter.

  In FIG. 1, an AWGN channel is used for the sake of brevity. Therefore, it is assumed that interference 26 and noise 27 are added to the signal output and transmitted from the perturbation vector adder 23 and received by the receiving device. Since it is an AWGN channel, the gain of the propagation path is 1.

  Next, the configuration of the receiving apparatus will be described in detail with reference to FIG. 1 and the flowchart of the receiving apparatus (FIG. 3).

  (Step S201) The dither sequence subtraction unit 33 subtracts the dither sequence added by the transmission device from the reception signal, and inputs the reception signal obtained by subtracting the dither sequence to the transmission bit soft estimation unit 35 and the perturbation bit soft estimation unit 37. . Further, when the transmission apparatus does not add the dither sequence, the dither sequence subtraction unit 33 is omitted.

  (Step S202) The transmission bit soft estimator 35 performs soft estimation considering the Modulo operation on the received signal obtained by subtracting the dither sequence, calculates a soft estimation value of the transmission bit, and inputs the soft estimation value to the Check Node 41. Hereinafter, as an example, a case will be described in which reception processing is performed using an LLR (Log Liklihood Ratio) as a soft estimation value of a transmission bit. Further, the receiving apparatus does not necessarily use the LLR as the soft estimation value in the reception process, and may use a likelihood ratio or the like. A soft estimated value of a transmission bit using LLR is called a transmission bit LLR. A method of calculating the transmission bit LLR in the transmission bit soft estimation unit 35 will be described with reference to FIGS.

In FIG. 5, ●, ▲, ○, and △ are candidate signal points corresponding to combinations of transmission bits and perturbation bits, respectively, described in FIG. FIG. 6 is an enlarged view of FIG. 5, and a received signal point ( _denoted as y _k . Here, the received signal is a received signal obtained by subtracting a dither sequence by the dither sequence subtracting unit 33) is shown as x. It is. First, the distances from the reception signal point to the four points closest to ●, ▲, ○, and △ are obtained.

候補信号点が●だったときに雑音の影響で受信信号点に移動する確率は、

となる。ここで、受信信号点の位置を変数xと置いた。また、同様に候補信号点が▲だったときに雑音の影響で受信信号点に移動する確率は、

であり、候補信号点が△だったときに雑音の影響で受信信号点に移動する確率は、

であり、候補信号点が○だったときに雑音の影響で受信信号点に移動する確率は、

When the candidate signal point is ●, the probability of moving to the received signal point due to noise is

It becomes. Here, the position of the reception signal point is set as a variable x. Similarly, when the candidate signal point is ▲, the probability of moving to the received signal point due to the noise is

When the candidate signal point is △, the probability of moving to the received signal point due to the influence of noise is

The probability of moving to the received signal point due to the noise when the candidate signal point is ○ is

となる。送信ビット軟推定部３５は、ステップＳ２０２においては、送信ビットLLR（Ltxk）を下式（１−７）で算出する。

It becomes. In step S202, the transmission bit soft estimator 35 calculates the transmission bit LLR (Ltxk) by the following equation (1-7).

  The transmission bit soft estimator 35 calculates the transmission bit LLR for the received signal based on the equation (1-7), and inputs it to the Check Node 41. The numerator of the part to which the log in Expression (1-7) is applied is “the distance from the reception signal point x to the point closest to the reception signal point x out of the signal points lined up for every Modulo width” and “each modulation line width”. The denominator of the portion to which the log in Equation (1-7) is a function of “the distance to the point closest to the received signal point x among the signal points ○” This is because it is a function of the “distance to the point closest to the received signal point x among the points ▲” and the “distance to the point closest to the received signal point x among the signal points Δ arranged for each Modulo width”. Therefore, it can be said that the transmission bit soft estimator 35 calculates the transmission bit by regarding the point translated by an integral multiple of the Modulo width as the same reception signal point.

  (Step S203) The perturbation bit soft estimator 37 performs perturbation bit soft estimation on the received signal obtained by subtracting the dither sequence, calculates the perturbation bit LLR, and inputs the perturbation bit LLR to the Check Node 41. The calculation method of the perturbation bit LLR in the perturbation bit soft estimation unit 37 is calculated by the following formula (1-8) using the formulas (1-3) to (1-6) similarly to the transmission bit soft estimation unit 35. .

  The perturbation bit soft estimator 37 calculates a perturbation bit LLR for the received signal based on Expression (1-8), and inputs it to the Check Node.

  As shown in FIGS. 5 and 6, the points ▲, ●, Δ, and ○ are periodically arranged every integer multiple (even multiples) of the Modulo width (predetermined size τ). It can be said that the bit soft estimator 37 calculates the perturbation bit by regarding the point translated by an integral multiple of 2 of the Modulo width as the same received signal point.

  (Step S204) Next, the Check Node 41 calculates a sign bit LLR from the transmission bit LLR and the perturbation bit LLR, and inputs it to the deinterleave unit 45.

  Now, each bit represented by bci, bpv, and btx has a relation represented by the following check matrix [I P I].

  Here, I is a unit matrix, and P is a 9 × 9 matrix indicated by P in Formula (1-1). The check matrix storage unit stores [I P I], and inputs the check matrix [I P I] to CheckNode when Check Node performs the operation of step S204. In step S204, Lci is calculated from Lpv and Ltx. As an example, a method for obtaining Lci4 will be described.

btx, bci, and bpv are subject to the nine constraints shown in equation (1-1), one of which is
0 = bci5 + btx5 + bpv1 + bpv3 (1-10)
It is expressed. In step S204, Ltx and Lpv are input to the Check Node, so Lci5 is obtained by using Ltx5, Lpv1, and Lpv3. This is because LLR is a value based on the probability that each bit is 0 or 1, so that Lci5 can be obtained by using that each bit always satisfies the constraint condition of Expression (1-10). A specific method for obtaining Lci5 will be described.

  According to the formula (1-10), the combination of (btx5, bpv1, bpv3) where bci5 = 1 is (0,0,1), (0,1,0), (1,0,0), (1, 1, 1), and bic5 = 0 in other combinations, so the probability pci5 (1) where bci5 is 1 and the probability pci5 (0) where 0 is 0 are calculated by the following equations.

によって求めることができる。これ以外にも同様にして、ptxk(1)とptxk(0)をLcikから、ppvk(1)とppvk(0)をLpvkから求めることができる。このように求めたpci5(1)とpci5(0)から、確率とLLRの変換式

を用いて、Lci5を算出することができる。また、ここで示した計算方式の代わりにLDPC符号等で用いられるSum-Product法を用いても算出することができる。式（１−１０）と同様の例を用いてSum-Productアルゴリズムを説明する（非特許文献５参照）。
まず、

とおく。ここでsign(x)は、x>0またはx=0で+1、x<0で-1となる関数である。
また関数φを

とおく。tanhはハイパボリック・タンジェントを表す。すると、Lci5は下式で求められる。 Here, ptx5 (1) and ptx5 (0) are from Ltx5,

Can be obtained. Similarly, ptxk (1) and ptxk (0) can be obtained from Lcik, and ppvk (1) and ppvk (0) can be obtained from Lpvk. From pci5 (1) and pci5 (0) obtained in this way, the conversion formula of probability and LLR

Can be used to calculate Lci5. The calculation can also be performed by using the Sum-Product method used in the LDPC code or the like instead of the calculation method shown here. The Sum-Product algorithm will be described using an example similar to Equation (1-10) (see Non-Patent Document 5).
First,

far. Here, sign (x) is a function that becomes +1 when x> 0 or x = 0, and −1 when x <0.
And the function φ

far. tanh represents a hyperbolic tangent. Then, Lci5 is obtained by the following formula.

という拘束があるとする。b_k以外の全てのビットのLLR（L_n）からb_kのLLR（L_k）を算出するとする。まず、

とおき、

と算出できる。 here,

Suppose that there is a constraint. Assume that LLR (L _k ) of b _k is calculated from LLR (L _n ) of all bits other than b _k . First,

Toki,

And can be calculated.

  (Step S205) The deinterleaving unit 45 deinterleaves the code bit LLR and inputs the deinterleaved code bit LLR to the decoding unit 53. Here, deinterleaving is to return the order of the code bits LLR to the order corresponding to the code bits before the interleaving section on the transmission side by performing a completely reverse replacement with the interleaving performed in the interleaving section 5 on the transmission side. To tell.

(Step S206) The decoding unit 53 performs decoding based on the code bit LLR, and then (Step S207) the number of iterations _nite stored in a memory (not shown) in the decoding unit 53 is a predetermined value. If the number _Nite has not been reached, the decoded code bit LLR is input to the code bit LLR subtraction unit 51, and if the predetermined number _Nite has been reached, the information bit LLR is output to the hard decision unit 55. Here, the information bit LLR represents the likelihood of the bit input to the encoding unit 3 on the transmission side.

  (Step S208) The code bit LLR subtraction unit 51 subtracts the code bit LLR before decoding from the code bit LLR after decoding, and inputs the subtracted code bit LLR to the interleaving unit 47.

  (Step S209) The interleaving unit 47 performs the same interleaving on the code bit LLR input from the code bit LLR subtracting unit 51 as the interleaving unit 47 on the transmission side, and inputs the interleaved code bit LLR to the Check Node 41. .

(Step S210) The Check Node 41 calculates the transmission bit LLR by the Sum-Product algorithm using the interleaved code bit LLR and the perturbation bit LLR input from the perturbation bit soft estimation unit 37, and the transmission bit soft estimation To the unit 35 and the perturbation bit soft estimation unit 37. In step S210, the perturbation bit LLR is repeated when the number of repetitions is 1 (n _ite = 1), and is repeated when the perturbation bit LLR calculated in step S203 is 2 or more (n _ite > 1). The perturbation bit LLR calculated in step S213 (described later) the number of times before is used.

  (Step S211) Using the code bit LLR interleaved by the Check Node 41, the perturbation bit LLR input from the perturbation bit soft estimation unit 37, and the transmission bit LLR input from the transmission bit soft estimation unit 35, the perturbation bit LLR is calculated. calculate.

  For example, when calculating the perturbation bit LLR Lpv3 using the equation (1-10) which is one of the equations (1-1), the code bit LLR Lci5 input from the interleave unit 47, and step S202 or Using the transmission bit LLR Ltx5 input from the transmission bit soft estimation unit 35 in step S212 (described later) and the perturbation bit LLR Lpv1 input from the perturbation bit soft estimation unit 37 in step S203 or step S213 (described later), The perturbation bit LLR Lpv3 is newly calculated by (1-10). At this time, the Sum-Product method may be used. At this time, the perturbation bit LLR Lpv3 input from the perturbation bit soft estimation unit 37 in step S203 or step S212 (described later) is not used, but is calculated from the LLRs of other bits included in the equation (1-10). .

  Similarly, when the perturbation bit LLR Lpv1 is calculated using the equation (1-10), the sign bit LLR Lci5 input from the interleave unit 47 and the transmission bit soft estimation unit in step S202 or step S212 (described later) Using the transmission bit LLR Ltx5 input from 35 and the perturbation bit LLR Lpv3 input from the perturbation bit soft estimator 37 in step S203 or step S213 (described later), a new perturbation bit LLR is obtained by equation (1-10). Calculate Lpv1.

In this way, the perturbation bit LLR is obtained for each bit included in each expression of the expression (1-1). However, in many cases, the same bits are included in a plurality of expressions as perturbation bits, such as bpv2 in the third and fourth lines of Expression (1-1). In this case, a value obtained by adding Lpv2 newly calculated by the constraint condition of the third row and Lpv2 newly calculated by the constraint condition of the fourth row is input to the transmission bit soft estimation unit 35 and the perturbation bit soft estimation unit 37. To do. In step S211, when the number of repetitions is 1 (n _ite = 1), the transmission bit LLR calculated in step S202 is used, and when the number of repetitions is 2 or more (n _ite > 1), the number of repetitions is 1. The transmission bit LLR calculated in the previous step S212 (described later) is used. Similarly, when the number of repetitions is 1 (n _ite = 1), the perturbation bit LLR calculated in step S203 is used, and when the number of repetitions is 2 or more (n _ite > 1), the number of repetitions is 1 time before. The perturbation bit LLR calculated in step S213 (described later) is used.

  (Step S212) Next, the transmission bit soft estimator 35 calculates a new transmission bit LLR from the perturbation bit LLR calculated by the Check Node 41 and the reception signal obtained by subtracting the dither sequence, and adds a new transmission bit LLR to the Check Node 41. Input the transmission bit LLR.

と算出できる。その後、ppvk(1)とppvk(0)と、式（１−３）から式（１−６）の値を用いて、 Specifically, the calculation is performed according to the following procedure. First, the probability that each bit of the kth Lpvk of the perturbation bit LLR calculated in step 211 is 0 and 1 is obtained. Similar to formula (1-12),

And can be calculated. Then, using the values of ppvk (1) and ppvk (0) and equations (1-3) to (1-6),

とLtxkを算出する。この方法によって、式（１−７）において算出した送信ビットLLRよりも情報量の高い送信ビットLLRを得ることができる。式（１−７）と比較して、誤り訂正符号により情報量が高められた摂動ビットLLRを用いて送信ビットLLRを算出しているからである。

And Ltxk is calculated. By this method, it is possible to obtain a transmission bit LLR having a higher information amount than the transmission bit LLR calculated in Expression (1-7). This is because the transmission bit LLR is calculated using the perturbation bit LLR whose amount of information is increased by the error correction code as compared with the equation (1-7).

  (Step S213) The transmission bit soft estimator 35 also calculates a new perturbation bit LLR from the transmission bit LLR calculated by the Check Node 41 and the received signal obtained by subtracting the dither sequence, and adds a new perturbation to the Check Node 41. Enter bit LLR.

と算出できる。その後、ptxk(1)、ptxk(0)、式（１−３）から式（１−６）の値を用いて、

によってLpvkを算出する。この方法によって、式（１−７）において算出した摂動ビットLLRよりも情報量の高い摂動ビットLLRを得ることができる。式（１−７）と比較して、誤り訂正符号により情報量が高められた送信ビットLLRを用いて摂動ビットLLRを算出しているからである。 Specifically, the calculation is performed according to the following procedure. First, the probability that each bit of the kth Ltxk of the transmission bit LLR calculated in step 210 is 0 and 1 is obtained. Similar to formula (1-12),

And can be calculated. After that, using the values of ptxk (1), ptxk (0), expressions (1-3) to (1-6),

To calculate Lpvk. By this method, it is possible to obtain a perturbation bit LLR having a higher information amount than the perturbation bit LLR calculated in Expression (1-7). This is because the perturbation bit LLR is calculated using the transmission bit LLR whose amount of information is increased by the error correction code as compared with the equation (1-7).

By repeating Step S204 to Step S213, the likelihood information represented by the transmission bit LLR, the perturbation bit LLR, and the sign bit LLR is enhanced. When the predetermined number of repetitions _Nite is finally reached, the decoding unit 53 outputs the information bit LLR to the hard decision unit 55 in step S207.

  (Step S214) Finally, the hard decision unit 55 makes a hard decision on the information bit LLR and outputs the decided information bit.

2. Modification 1 (MU-MIMO THP 2MTs)
Next, a method for applying the DPC encoding unit and the DPC decoding unit to the downlink MU-MIMO THP will be described. This modification is a system in which two terminals each having one antenna are spatially multiplexed by a base station having two antennas and signals are transmitted on the downlink (that is, transmission signals addressed to two receivers simultaneously at the same frequency) The system of the above-described element part is applied to a system that transmits

2.1: Base station apparatus The structure of the base station apparatus in this modification is demonstrated using FIG. The base station apparatus B includes an antenna 101a / b, a base station transmission unit 103, a base station reception unit 105, a linear filter calculation unit 151, an encoding unit 153, a modulation unit 155, an interference calculation unit 157, and a DRS (Dedicated Reference Signal) generation unit. 161, an eigensignal configuration unit 163a / b, a linear filter multiplication unit 165, and a DPC encoding unit 167.

  The base station transmission unit 103 includes radio transmission units 111a and 111b, GI (Guard Interval) insertion units 113a and b, IFFT (Inverse Fast Fourier Transform) units 115a and b, and a frame configuration unit. 117a · b and a CRS (Common Reference Signal) generator 121.

  The base station receiving unit 105 includes a radio receiving unit 131a / b, a GI removing unit 133a / b, an FFT (Fast Fourier Transform) unit 135a / b, and a propagation path state information acquiring unit 137. In addition, the DPC encoding unit 167 may use the same configuration as that described above (DPC encoding unit 1 in FIG. 1).

  First, the operation of the base station receiving unit 105 will be described. The radio receivers 131a and 131b receive signals including propagation path state information transmitted from the terminals via the antennas 101a and 101b, downconvert them to baseband, perform analog / digital conversion, and perform the digital The signals are input to the GI removal units 133a and 133b, respectively. Each GI removal unit 133a · b removes the GI from the input digital signal and inputs the GI to the FFT unit 135a · b. Each FFT unit 135a / b inputs the frequency domain signal calculated by performing FFT on the input signal to the propagation path state information acquisition unit 137. The propagation path state information acquisition unit 137 acquires propagation path state information (Channel State Information: CSI) from the input frequency domain signal and inputs it to the linear filter calculation unit 151. Here, the propagation path state information is a value indicating the complex gain of the propagation path from each base station antenna to the antenna of each terminal, and can be expressed by the following equation.

  The pth and qth column components (p, q = 1, 2) of this matrix represent the complex gain of the propagation path between the pth terminal and the qth antenna of the base station. Also, this propagation path state information is acquired by the base station for each subcarrier.

  FIG. 7 illustrates the case where the propagation path state information is acquired from two antennas of the base station as an example, but the propagation path state information may be acquired using a signal from only one antenna. . In FIG. 7, the base station receiving unit 105 performs OFDM (Orthogonal Frequency Division Multiplexing) using the GI removing units 133a and 133b and the FFT units 135a and 135b. The channel state information may be acquired by other transmission methods such as MC-CDM (Multi-Carrier Code Division Multiplexing). Further, although it has been described that the propagation path state information is acquired by the base station for each subcarrier, it may be propagation path state information for a plurality of subcarriers.

  Next, the linear filter calculation unit 151 that has acquired the propagation path state information calculates a linear filter and an interference coefficient filter by the following QR decomposition.

Here, ^H represents Hermitian conjugate. QR decomposition is an operation that decomposes an arbitrary matrix so that it is represented by the product of a unitary matrix Q and an upper triangular matrix R. The linear filter is Q, and the interference coefficient filter F is

と算出できる。線形フィルタ算出部１５１は、線形フィルタQを線形フィルタ乗算部１６５に入力し、干渉係数フィルタFを干渉算出部１５７に入力する。

And can be calculated. The linear filter calculation unit 151 inputs the linear filter Q to the linear filter multiplication unit 165 and inputs the interference coefficient filter F to the interference calculation unit 157.

によって干渉gを算出する。算出した干渉gの実部をDPC符号部１６７に入力する。本実施例では、I-chにのみデータ信号が乗るBPSK方式を用いているので、DPC符号部１６７に入力するのは干渉gの実部のみである。DPC符号部１６７は、前述の方法によって２番目の端末宛のデータ信号（前述の例においては「DPC符号後信号」と呼んでいたもの。）を算出する。ここで、図１において、干渉f_kと表していたものが、干渉算出部１５７で算出した干渉gに対応している。各OFDMシンボルの各サブキャリアにおける干渉gを実部及び虚部の２つに分解したものが干渉f_kとなる。前述のようにDPC符号部１６７において第２の端末宛のデータ信号を算出した後、DRS生成部１６１は固有信号構成部１６３ｂに第２の端末宛の送信信号を入力する。ここで、DRS生成部１６１は、２つの固有信号構成部１６３ａ・ｂにDRSを入力する。第１の固有信号構成部１６３ａは、第１の端末宛の送信信号にDRSを挿入する。また、第２の固有信号構成部１６３ｂは、第２の端末宛の送信信号に対してDRSを挿入する。 Encoding section 153 performs error correction coding on the information bits addressed to the first terminal, and inputs the post-coding bits addressed to the first terminal to modulation section 155. Modulation section 155 modulates the code bit to generate modulated signal w ₁ addressed to the first terminal, and inputs it to specific signal configuration section 163a and interference calculation section 157. Modulation section 155 inputs a modulated signal addressed to the first terminal (also referred to as a data signal addressed to the first terminal) to interference multiplication section 157. In addition, the interference calculation unit 157 calculates interference with respect to the modulation signal addressed to the second terminal using the modulation signal addressed to the first terminal. The interference calculation unit 157

To calculate the interference g. The calculated real part of interference g is input to DPC encoding section 167. In the present embodiment, since the BPSK method in which the data signal is carried only on I-ch is used, only the real part of interference g is input to the DPC encoding unit 167. The DPC encoding unit 167 calculates the data signal addressed to the second terminal (which was called “signal after DPC encoding” in the above example) by the above-described method. Here, what is represented as interference f _k in FIG. 1 corresponds to the interference g calculated by the interference calculation unit 157. The interference f _k is obtained by decomposing the interference g in each subcarrier of each OFDM symbol into two parts, a real part and an imaginary part. As described above, after the DPC encoding unit 167 calculates the data signal addressed to the second terminal, the DRS generation unit 161 inputs the transmission signal addressed to the second terminal to the unique signal configuration unit 163b. Here, the DRS generation unit 161 inputs the DRS to the two unique signal configuration units 163a and 163b. The first unique signal configuration unit 163a inserts the DRS into the transmission signal addressed to the first terminal. In addition, second unique signal configuration section 163b inserts a DRS into the transmission signal addressed to the second terminal.

  Thereafter, both eigensignal forming sections 163a and 163b each input the transmission signal into which the DRS is inserted to the linear filter multiplying section 165. The linear filter multiplication unit 165 inputs the transmission signal for the first antenna calculated by multiplying the input signal by the linear filter Q to the first frame configuration unit 117a. Similarly, the linear filter multiplication unit 165 inputs the transmission signal for the second antenna 101b to the second frame configuration unit 117b.

  Finally, the operation of the base station transmission unit 103 will be described. The CRS generation unit 121 in the base station transmission unit 103 inputs the CRS to the first and second frame configuration units 117a and 117b. Each frame configuration section 117a and 117b configures a frame by 1) inserting a CRS into the transmission signal for each antenna input from the linear filter multiplication section 165. Each of the frame configuration units 117a and 117b includes 2) a frame including CRS but not including a transmission signal for each antenna input from the linear filter multiplication unit 165, and 3) including no CRS but including a linear filter multiplication unit 165. A frame including a transmission signal for each antenna input from may be configured, and the two types of frames may be transmitted as different frames. The frame configuration units 117a and 117b input the configured frames to the first and second IFFT units 115a and 115b, respectively. Each IFFT unit 115a / 115b performs IFFT on the input frame, calculates a time domain signal, and inputs it to the first and second GI insertion units 113a / 113b, respectively. Each of the GI insertion units 113a and 113b inserts a GI into the input time domain signal and inputs the GI into the first and second wireless transmission units 111a and 111b. Each of the wireless transmission units 111a and 111b performs digital / analog conversion on the time domain signal in which the GI is inserted, and then transmits the wireless signal generated by up-conversion to the carrier frequency via the corresponding antenna.

2.2 Terminal Device Next, the configuration of the receiving device in this modification will be described. In this modification, there are a terminal (referred to as a first terminal) that receives a signal that is not DPC-encoded and a terminal (referred to as a second terminal) that receives a signal that is DPC-encoded. So I will explain both.

  First, the configuration of a terminal (first terminal) that receives a signal that is not DPC encoded will be described with reference to FIG. The terminal apparatus includes an antenna 201, a radio reception unit 203, a GI removal unit 205, an FFT unit 207, a signal separation unit 211, a propagation path estimation unit 215, a propagation path state information generation unit 217, an IFFT unit 221, a GI insertion unit 223, A radio transmission unit 225, a propagation path compensation unit 227, a demodulation unit 231, and a decoding unit 233 are included.

  The wireless reception unit 203 receives a signal transmitted from the base station via the antenna 201, down-converts the signal to baseband, performs analog / digital conversion, converts the signal into a digital signal, and removes the digital signal from the GI Force the part 205. The GI removal unit 205 removes the GI from the input digital signal and inputs it to the FFT unit 207. Each FFT unit 207 inputs a frequency domain signal calculated by performing FFT on the input signal to the signal separation unit 211. The signal separation unit 211 separates CRS and DRS from the input frequency domain signal and inputs them to the propagation path estimation unit 215. Similarly, the data signal addressed to the first terminal is separated from the frequency domain signal and input to the propagation path compensation unit 227.

  The propagation path estimation unit 215 performs propagation path estimation based on the CRS, and inputs information indicating the propagation path state to the propagation path state information generation unit 217. The propagation path state information generation unit 217 generates a signal including the propagation path state information based on the information indicating the propagation path state input from the propagation path estimation unit 215, and inputs the signal to the IFFT unit 221. Here, in principle, the propagation path state information is preferably a complex gain of a propagation path between each antenna of the base station and the antenna of the first terminal, but even if the complex gain is quantized. It may be a codebook number shared with the base station apparatus, or may be a method of transmitting channel state information after being compressed.

  IFFT section 221 performs IFFT on the signal including the propagation path state information to calculate a time domain signal, and inputs the time domain signal to GI insertion section 223. The GI insertion unit 223 inserts a GI into the time domain signal and inputs it to the wireless transmission unit 225. The radio transmission unit 225 performs digital / analog conversion on the time domain signal in which the GI is inserted, and then transmits the radio signal generated by up-conversion to the carrier frequency to the base station apparatus via the antenna.

  Further, the propagation path estimation unit 215 is based on the DRS and can be regarded as a propagation path in which the data signal addressed to the first terminal has substantially passed, that is, the linear filter multiplication unit 165 (FIG. 7) of the base station apparatus and the actual The complex gain of the equivalent propagation path including the propagation path is estimated and input to the propagation path compensation unit 227. The propagation path compensation unit 227 performs propagation path compensation on the data signal addressed to the first terminal based on the complex gain input from the propagation path estimation unit 215, and sends the modulated signal after propagation path compensation to the demodulation unit 231. input. The demodulator 231 inputs the hard decision value or soft estimate obtained by demodulating the modulated signal to the decoder 233. The decoding unit 233 performs decoding using the input hard decision value or soft estimation value, and extracts information bits.

  Next, the configuration of a terminal (second terminal) that receives a DPC-encoded signal will be described with reference to FIG. The configuration of the second terminal has a configuration in which the demodulating unit 231 and the decoding unit 233 of the first terminal are replaced with the DPC decoding unit 235, and the operations other than the DPC decoding unit 235 correspond to the first terminal. It is the same operation as the part to do. Further, the DPC decoding unit 235 performs the operation as described above with reference to FIG. 1 and the like and outputs information bits. In the above description, the input signal of the DPC decoding unit called the received signal is treated as a signal after propagation path compensation input from the propagation path compensation unit in this modification.

3. Modification 2 (MU-MIMO THP N-MTs)
In the first modification, two terminals are multiplexed. In the second modification, N terminals are multiplexed (that is, a system that simultaneously transmits transmission signals addressed to a plurality of receiving devices at the same frequency). Will be described.

3.1 Base Station Apparatus A configuration example of the base station apparatus in this modification will be described with reference to FIG. In the first modification, the base station apparatus has two antennas, but in this modification, as shown in FIG. 10, the base station apparatus has N (301-1 to 301-N). The base station transmission unit 303 and the base station reception unit 305 also include radio transmission units 311-1 to N, GI insertion units 313-1 to N, IFFT units 3151 to N, frame configuration units 317-1 to N, and radio reception units 331. -1 to N, GI removal units 333-1 to 333 -N, and FFT units 335-1 to 335 -N are provided in the same number as the number N of antennas. Since these components perform operations corresponding to the respective antennas and the respective operations are the same as those of the first modification, the description thereof is omitted here. However, the propagation path state input to the linear filter calculation unit 351 from the propagation path state information corresponding to the N antennas and the number of terminals N is N × N expressed by the following equation. It becomes a matrix.

  The pth and qth column components (p, q = 1, 2,..., N) of this matrix represent the complex gain of the propagation path between the pth terminal and the qth of the base station. Next, the linear filter calculation unit 351 that acquired the propagation path state information calculates a linear filter and an interference coefficient filter using QR decomposition as in the first modification. At this time, the linear filter calculation unit 351 can also use the above equations (1-102) and (1-103) in this modification, and the linear filter Q calculated by both equations is used as the linear filter multiplication unit 365. The interference coefficient filter F is input to the interference calculation unit 357.

によって干渉ｆを算出する。ここで、wは、あるサブキャリアにおける第１番目の端末宛の変調信号及び第２〜Ｎ番目の端末宛のDPC符号後信号を第１成分から第Ｎ成分によって表した縦ベクトルである。また、ｆは、wと同じサブキャリアにおける第１〜Ｎ番目の端末宛が受ける干渉信号を第１成分から第Ｎ成分によって表した縦ベクトルである。１番目の端末宛の変調信号w₁を算出した後は、干渉件数フィルタFが下三角行列であることを用いて下記手順によって順番にｆとwとを算出していく。
１）干渉算出部３５７は、第２の端末の干渉信号ｆ_２を、式（１−１０６）の第２列のみを取り出した、

として第２の端末に対応するDPC符号部３６７−２に干渉信号g_２の実部を入力する。なお、[ ]_kはベクトルの第k成分のみを取り出す演算を表している。本実施例では、実部にのみにビットを配置するBPSK方式を用いているので、DPC符号部３６７−２に入力するのは信号g_２の実部のみとする。 The encoding unit 353 performs error correction coding on the information bits addressed to the first terminal, and inputs the post-coding bits addressed to the first terminal to the modulation unit 355. Modulation section 355 modulates the post-code bits to generate modulated signal s ₁ addressed to the first terminal, and inputs it to specific signal configuration section 363-1 and interference calculation section 357. Modulation section 355 inputs a modulated signal addressed to the first terminal (also referred to as a data signal addressed to the first terminal) to linear filter multiplication section 365. Further, the interference calculation unit 357 calculates interference with respect to the modulation signal addressed to the second terminal using the modulation signal addressed to the first terminal. The interference calculation unit 357

To calculate the interference f. Here, w is a vertical vector in which a modulated signal addressed to the first terminal and a post-DPC code signal addressed to the 2nd to Nth terminals in a certain subcarrier are represented by the first to Nth components. Further, f is a vertical vector in which interference signals received by the first to Nth terminals in the same subcarrier as w are represented by the first component to the Nth component. After calculating the modulation signal w ₁ addressed to the first terminal, f and w are calculated in order according to the following procedure using the fact that the interference number filter F is a lower triangular matrix.
1) The interference calculation unit 357 extracts only the second column of the expression (1-106) from the interference signal f2 of the _second terminal.

As you enter the real part of the DPC coding unit 367-2 in the interference signal g ₂ corresponding to the second terminal. [] _K represents an operation for extracting only the k-th component of the vector. In this embodiment, because of the use of BPSK method of placing the bits only in the real part, to input to the DPC coding unit 367-2 is the only real part of the signal g _2.

2) the DPC coding unit 367-2 corresponding to the second terminal calculates a data signal w ₂ of the second to the terminal based on the interference signal g ₂ of the second information bits and the second terminal to the terminal The unique signal configuration unit 363-2 and the interference calculation unit 357 are input.

によって算出し、第２の端末に対応するDPC符号部３６７−２に干渉信号g_３の実部を入力する。 3) The interference calculation unit outputs the interference signal g ₃ of the third terminal for each subcarrier.

Calculated by inputting the real part of the interference signal g ₃ to the DPC coding unit 367-2 corresponding to the second terminal.

4) the DPC coding unit 367-3 is DPC code after signal w ₃ of the third to the terminal based on the interference signal g ₃ of the information bit of the addressed third terminal third terminal corresponding to the third terminal It is calculated and input to the unique signal configuration unit 363-3 and the interference calculation unit 357.

Thereafter, the data signal is sequentially calculated for each terminal in the same procedure as 1) to 4). That is, after the interference calculation unit 357 calculates the interference signal g _k addressed to the k-th terminal according to the equation (1-106), the information addressed to the k-th terminal in the corresponding DPC encoding unit of the k-th terminal. The data signal w _k addressed to the k-th terminal is calculated using the bit and the real part of the interference signal g _k , and input to the specific signal configuration unit 363-k and the interference calculation unit 357. This operation is repeated until the calculated N-th data signal w _N destined for the terminal.

The operations of the DRS generation unit 361, the specific signal configuration unit 363, the linear filter multiplication unit 365, and the base station transmission unit 303 are the same as those in the first modification, and thus description thereof is omitted here.
The above is the operation of the base station apparatus in this modification.

3.2 Terminal Device This modification includes a terminal (first terminal) that receives a signal that is not DPC-encoded as described in Modification 1 above, and a terminal that receives a signal that is DPC-encoded (second terminal). Use a terminal that operates exactly the same as the terminal. The first terminal is a terminal that receives a signal that is not DPC-encoded as in the first modification, and the second to Nth terminals are the same as the terminals that receive the DPC-encoded signal in the first modification. It has a configuration (FIG. 9). In the present embodiment, Modulo Loss, which has been a problem in the MU-MIMO THP scheme, can be reduced using the error correction code capability.

  In this modified example, the linear filter and the interference coefficient filter are calculated by a method using QR decomposition of ZF (Zero-Forcing) type generally used in downlink MU-MIMO THP. As described, ordering may be used for QR decomposition, and a linear filter and an interference coefficient filter may be calculated based on a MMSE (Minimum Mean Square Error) standard. Further, as described in Non-Patent Document 3, the linear filter and the interference coefficient filter may be calculated using the BLAST method based on ZF or MMSE. Also, although it has a plurality of reception antennas physically, it is assumed that a terminal that performs demodulation and decoding processing assuming that one reception signal is substantially received by synthesizing reception signals by each antenna. The technology described in this modification can be applied as a terminal having one antenna described in the modification.

<Second Embodiment>
In the first embodiment, the case where the modulation scheme is BPSK has been described, but in this embodiment, the case where the modulation scheme is QPSK (Quadrature Phase Shift Keying) will be described. As in the first embodiment, in this embodiment, first, the operation of the element part of the invention will be described using an AWGN channel, and then a modification in which the element part is applied to the downlink MU-MIMO THP. explain.

  First, the operation of the DPC encoding unit that is an element part of the embodiment of the present invention will be described. FIG. 11 shows a configuration example of a DPC encoding unit 401 and a DPC decoding unit 451 which are element parts of this embodiment. The DPC encoding unit 401 in the present embodiment adds a real part / imaginary part separation unit 411 and a real part / imaginary part combining unit 415 to the DPC encoding unit 1 (FIG. 1) according to the first embodiment. It has a configuration. In addition, the interference signal input to the DPC encoding unit 401 has a complex value as opposed to a real value in the first embodiment.

  The DPC encoding unit 401 according to the present embodiment calculates a data signal by dividing a real part (I-ch) and an imaginary part (Q-ch), and later, a real part / imaginary part combining unit 415 It has a configuration for generating a signal with information on the imaginary part. The modulation unit 17 in the DPC encoding unit 401 performs modulation for each bit of the transmission bits btxk. Normally, QPSK modulates 2 input bits and outputs one modulation symbol. In the DPC encoding unit 401, a signal having a component only in one real part per 1 bit as in the BPSK system. Output. Finally, since the synthesis is performed by the real part / imaginary part synthesis unit, substantially the same operation as in the case of BPSK modulation can be performed.

として出力する。ここでａは任意の自然数を示す。実部・虚部合成部４１５は、実部及び虚部に成分を持つデータ信号ｘ_ｌを送信する。 Also, the real and imaginary part separating section 411 separates the interference f _l complex value input to DPC coding portion 401 to the real part and the imaginary part, an input to the interference subtraction unit 19 as an interference signal f _k alternately To do. Other operations other than the real part / imaginary part combining unit 415 are the same as those in the first embodiment. A data signal x _k composed of only real part signals is calculated by the same method as in the first embodiment. Thereafter, in the real and imaginary part combining unit 415 arranges the data signals x _k with two by two sets in a real part and an imaginary part. That is,

Output as. Here, a represents an arbitrary natural number. Real and imaginary part combining unit 415 transmits the data signals x _l having a component in the real part and the imaginary part.

を実部のみにおける雑音分散とおいており、第２の実施の形態でも、dither系列減算部３３以降は全て信号を実数として扱って信号を算出するため、また式（１−３）から式（１−６）と同様の式を用いることができる。 The DPC decoding unit 451 in the present embodiment has a configuration in which a real part / imaginary part separation unit 453 is added before the dither sequence subtraction unit of the DPC decoding unit of the first embodiment. When the received signal Y ₁ is taken, the real / imaginary part separating unit 453 generates a received signal y _{k in} which Y ₁ is alternately arranged into a real part and an imaginary part, and inputs the received signal y _k to the dither sequence subtracting unit 33. Further, the dither sequence subtracting unit 33 and the subsequent operations are the same as those in the first embodiment. Note that in the equations (1-3) to (1-6) in the first embodiment, the noise variance is

In the second embodiment, since the dither sequence subtracting unit 33 and after all treat the signal as a real number and calculate the signal in the second embodiment, the equations (1-3) to (1 The same formula as −6) can be used.

  The above is the operation of the element portion of the present invention when QPSK is used. Also in the second embodiment using QPSK, the base station apparatus has the configuration shown in FIGS. 7 and 10 as in the first and second modifications of the first embodiment. However, in the first embodiment, only the real part of the interference signal calculated by the interference calculation unit 157 is input to the DPC encoding unit 1, but in the first embodiment, the real part and the imaginary part are combined. The combined complex number is input to the DPC encoding unit 401. This is because, in the second embodiment, the real / imaginary part separating unit 411 in the DPC encoding unit 401 separates the interference component into a real part and an imaginary part and uses them.

  In this embodiment, the invention according to the first embodiment is applied to QPSK, and Modulo performs communication at twice the rate as compared with the first embodiment based on BPSK. Loss can be reduced.

(Third embodiment)
In the first embodiment, the case where the modulation scheme is BPSK and the second embodiment 2 has been described in the case where the modulation scheme is QPSK. However, in this embodiment, the modulation scheme is Gray Mapping (gray mapping). A case of a method capable of demodulating the bits arranged in one modulation signal as independent, such as 16QAM (Quadrature Amplitude Modulation) using 64 and 64QAM will be described. Hereinafter, a is the number of bits arranged in one modulation signal. Also, l represents the modulation signal number.

1. Regarding the DPC code part In the second embodiment, the bits are calculated by dividing the real part and the imaginary part of one modulation symbol, and the code bits, transmission bits, and perturbation bits correspond one-to-one. . In the present embodiment, a transmission signal is calculated for each modulation symbol that combines the real part and the imaginary part.

  Therefore, in this embodiment, a code symbol, a transmission bit, and two perturbation bits for the real axis and the imaginary axis correspond to one modulation symbol.

  For example, in the case of 16QAM, one modulation symbol corresponds to four code bits and two transmission bits and two perturbation bits. In the case of 64QAM, one code symbol and six transmission bits correspond to one modulation symbol. And two perturbation bits.

  The operation of the DPC encoding unit of the present embodiment will be described in order with reference to the configuration diagram of FIG. 12 and the flowchart of FIG.

(Step S305) interference subtraction unit 19 subtracts the interference signal f _l from l-th modulation signal.

(Step S306) dither sequence adding section 21 adds the dither sequence d _l to l-th modulation signal after interference subtraction.

(Step S307) perturbation vector adder 23, the modulated signal after addition of l-th dither sequence, by adding the perturbation vector to generate the DPC code after signal X _l. The addition of the perturbation vector corresponds to performing the calculation of only the real part in the first embodiment on both the real part and the imaginary part. In addition, a perturbation vector z _2l−1 + jz _2l represented by a complex number is input to the perturbation bit calculation unit 25.

(Step S309) Finally, the post-DPC code signal _Xl is transmitted.

2. About the DPC decoding part of 3rd Embodiment In the 3rd Embodiment of this invention, in the DPC encoding part, since the signal of a real part and an imaginary part was calculated simultaneously, also in DPC decoding, the 1st Processing different from that in the embodiment and the second embodiment is performed. As later, in the description of this embodiment, the data signal output from the DPC coding unit 15, is denoted by X _I.

である。ここで、p(A)を事象Aが発生する確率とする。 First, the operation of the transmission bit soft estimation unit 35 will be described. When l of (El) th data signal received at the terminal now, the received signal is assumed to be Y _l. Here, the reception signal Y ₁ is a reception signal obtained by subtracting the dither sequence by the dither sequence subtraction unit 33. The transmission bits assigned to the l-th data signal are d_1, ..., d_a, the perturbation bits corresponding to the imaginary axis are d_a + 1, and the perturbation bits corresponding to the real axis are d_a + 2 (d_1, ..., d_a + 2 Are all “0” or “1”). D_k may be expressed as d _k (k is an arbitrary natural number). Then, the probability that the transmission bit arranged in the l-th data signal is received as the reception signal Y ₁ due to the influence of noise is as follows:

It is. Here, let p (A) be the probability that event A will occur.

Using the above equation (3-1), ^a total of 2 ^{a + 2} values are calculated when (d_1,..., D_a + 2) takes “0” and “1”, respectively. Thereafter, the transmission bit soft estimation unit 35 outputs a soft estimation value of each transmission bit assigned to the real axis in the following equation.

Here, “D” represents all 2 ^{a + 2} values that (d_1,..., D_a + 2) can take, and “D | d_c = 0” can be 2 that (d_1,..., D_a + 2) can take. ^{Among the a + 2} values, 2 ^{a + 1} values in which the c-th component d_c is 0 are shown.

In the above formula (3-2), the sum of the values of the formula (3-1) for 2 ^{a + 1} pieces of the numerator where the logarithm log is applied and the c-th component d_c is 0 is obtained, and the c-th component is obtained by the denominator. The sum of the values of Expression (3-1) for 2 ^{a + 1} pieces where d_c is 1 is taken.

とおいて、式（３−２）を算出する。２回目以降は、CheckNodeから入力された、摂動ビットLLRに従って、

を算出する。送信ビット軟推定部は、上記式（３−２）で算出した送信ビットLLRを算出した後、当該送信ビットLLRをCheck Nodeに入力する。 In the above equation (3-2), the same equation can be used in the first and subsequent iterations (corresponding to step S202 in the first embodiment) and the second (step S212 in the first embodiment) of the DPC decoding unit. . However, at the first time, information regarding the probability of taking each transmission bit is not input from Check Node,

Then, formula (3-2) is calculated. From the second time on, according to the perturbation bit LLR input from CheckNode,

Is calculated. The transmission bit soft estimator calculates the transmission bit LLR calculated by the above equation (3-2), and then inputs the transmission bit LLR to the Check Node.

  Next, the operation of the perturbation bit soft estimation unit 37 will be described. The perturbation bit soft estimation unit 37 also performs substantially the same operation as the transmission bit soft estimation unit 35. The perturbation bit soft estimator 37 also calculates each perturbation bit LLR by the following formula using the probability calculated by formula (3-1).

を算出する。 Here, e has a value of 1 or 2. In the expression (3-4), the same applies to the first and second iterations (corresponding to step S203 of the first embodiment) and subsequent detections (corresponding to step S213 of the first embodiment) of the DPC decoding unit. Equations can be used. However, since the information regarding the probability of taking each perturbation bit is not input from the Check Node at the first time, the probability that all the transmission bits are “0” or “1” is set to 1/2, respectively, and Expression (3-4) ) To calculate the perturbation bit LLR. From the second time on, according to the perturbation bit LLR input from CheckNode,

Is calculated.

  After calculating the perturbation bit LLR calculated by the above equation (3-4), the perturbation bit soft estimation unit 37 inputs the perturbation bit LLR to the Check Node.

2. Explanation when Example 3 is applied to MU-MIMO THP This embodiment is achieved by replacing the DPC encoding unit and DPC decoding unit of the same MU-MIMO THP system as in Example 2 with those described in this example. An example can be applied to the MU-MIMO THP system.

  According to the present embodiment, the invention described in the first embodiment and the second embodiment can be applied to 16QAM, 64QAM, etc., and high-rate communication can be performed while reducing Modulo Loss. can do. Further, although this embodiment is premised on a modulation system having a high multi-level number such as 16QAM, it can also be applied to QPSK.

<Fourth embodiment>
From the first embodiment to the third embodiment, interference received by each terminal in the interference subtraction unit 26 is completely removed. However, in this embodiment, the interference is intentionally removed incompletely. An example of applying ILP (Inflated Lattice Precoding) that can use residual interference to increase signal power to the present invention will be described.

  One configuration example of the communication apparatus according to the present embodiment will be described with reference to FIG. FIG. 16 has a configuration in which coefficient multiplication units 603 and 653 are newly provided in the DPC encoding unit 601 and the DPC decoding unit 651, respectively, with respect to FIG. 12 which is the configuration example of the third embodiment. The DPC encoding unit 601 is provided with a coefficient multiplying unit 603 before the interference signal is input to the real / imaginary part separating unit 411, and the DPC decoding unit 651 is configured to input the received signal to the dither sequence adding unit 33. A coefficient multiplier 653 is provided.

と表すことができる（係数αは１未満の正の係数である）。

(The coefficient α is a positive coefficient less than 1).

Next, a method of applying the communication technique according to the present embodiment to MU-MIMO THP will be described using the functional block diagram shown in FIG. In FIG. 17, in addition to the configuration of FIG. 10, a coefficient calculation unit 701 is newly provided in the base station apparatus. In addition to the equations (1-102) and (1-103), the linear filter calculation unit 351 uses the diagonal component of ^RH , which is an equivalent channel combining the linear filter Q and the actual channel H, as a coefficient. Input to the calculation unit 701. The coefficient calculation unit calculates the coefficient α calculated from the equation (4-1) using the diagonal component of ^{RH for} the second and subsequent terminals, and the DPC encoding units 367-2 to 367-2 of the corresponding terminals. Enter in N. In addition, the signal including information indicating the coefficient α is input to the specific signal configuration sections 363-1 to 363 -N of the corresponding terminals.

  The unique signal configuration units 363-2 to 363 -N constituting the unique signal addressed to the second and subsequent terminals use the data signal and the signal including the information indicating the coefficient α in addition to the DRS to generate the unique signal addressed to each terminal. Constitute.

  A configuration example of the terminal is shown in FIG. The configuration shown in FIG. 18 is a configuration example of second and subsequent terminals that perform DPC decoding. Since the first terminal that does not perform DPC decoding is not notified of the coefficient α, the same configuration as in FIG.

  In FIG. 18, a coefficient detection unit 801 is newly added to the configuration. When the signal separation unit 211 separates the unique signal addressed to each terminal, in the present embodiment, since the signal including the information indicating the coefficient α is newly included, the signal including the information indicating the coefficient α is included in the coefficient detection unit. 801 is input. Thereafter, the coefficient detection unit 801 detects the coefficient α and inputs it to the DPC decoding unit 235.

  With the above configuration, the ILP can be applied to the present invention, and the characteristics can be improved as compared with the communication techniques according to the first to third embodiments.

<Fifth embodiment>
In the MU-MIMO THP system described in the first to fourth embodiments, the perturbation vector is individually calculated for each terminal. In the fifth embodiment of the present invention, an example will be described in which VP is applied to improve power efficiency in the entire base station apparatus by optimizing in consideration of perturbation vectors of all terminals.

  Since VP is a technique premised on being applied to MU-MIMO, the base station apparatus has N transmission antennas and multiplexes N terminals, as in Modification 2 of the first embodiment. The case will be described. First, the configuration of the base station apparatus will be described using FIG. 19 and FIG. Unlike the first embodiment, in the communication technique according to this embodiment, DPC encoding is performed on signals addressed to all terminals. Therefore, in FIG. 19, the information bits of the first terminal are also input to the DPC encoding unit 367-1. In FIG. 20, the interference subtraction unit 19 that subtracts the interference signal from the modulation signal in FIG. 1 is eliminated, and the modulation unit 17 inputs the modulation signal directly to the dither sequence addition unit 21. In FIG. 19, each DPC encoding unit 367-1 to 367 -N outputs a signal after DPC encoding (a signal output from the perturbation vector addition unit 23 in FIG. 20) and inputs it to the perturbation vector calculation unit 903. . The perturbation vector calculation unit 903 uses the linear filter input from the linear filter calculation unit 351 and the post-DPC code signals input from the DPC encoding units 367-1 to 367-N, and adds the perturbation vector to each post-DPC code signal. And the perturbation vectors are respectively input to the corresponding DPC encoding units 367-1 to 367 -N.

  Here, the linear filter calculation unit 351 calculates a linear filter different from those in the first to fourth embodiments. There are ZF and MMSE types of linear filters, which can be expressed as follows using the propagation path state H, respectively.

Here, v = tr (R _μ ) using transmission power E _TR per sub-carrier 1 OFDM symbol and a diagonal matrix R _μ having a variance value of the n th noise in the n th row and the n th column. / E _tr . Here, tr () represents an operation for taking a matrix trace. The linear filter calculation unit 351 inputs the linear filter represented by the equation (5-1) or the equation (5-2) to the linear filter multiplication unit 365 and the perturbation vector calculation unit 903.

  Next, the operation of the perturbation vector calculation unit 903 will be described. Now, the post-DPC code signal of each terminal input to the perturbation vector calculation unit is represented by a vector u having each component. A perturbation vector calculation unit 903 calculates an optimal perturbation vector z to be added to the real part and the imaginary part of each component of the vector u.

を最小にするModulo幅τの整数倍で表される摂動ベクトルである。 Here, the optimal perturbation vector is

Is a perturbation vector represented by an integral multiple of the Modulo width τ.

を満たす。この最適なベクトルｚを見つける方法は、非特許文献４に記載されているSphere Encoding等によって実現できる。

Meet. This method of finding the optimal vector z can be realized by Sphere Encoding or the like described in Non-Patent Document 4.

  The perturbation vector calculation unit 903 inputs the component of the optimized perturbation vector for each terminal to the corresponding DPC encoding unit 367 of each terminal. The DPC encoding unit 367 inputs the perturbation vector input from the perturbation vector calculation unit 903 to the perturbation vector addition unit 23 and outputs a signal after DPC encoding. .

  In this method, the perturbation vector is calculated by the perturbation vector calculation unit 903 at the same time for both the real part and the imaginary part. Therefore, each perturbation bit calculation unit 903 calculates a perturbation bit by 2 bits. The perturbation bit adder also calculates transmission bits using a generator matrix corresponding to the input of perturbation bits every 2 bits. For example, QPSK in which one transmission bit is arranged in the real part and the imaginary part has a generation matrix in which two transmission bits and perturbation bits are combined corresponding to the real part and the imaginary part as shown in FIG. That is, in the first embodiment or the like, 1 is included in the lower triangular component excluding the diagonal component. However, in this embodiment, 2 × as shown by the solid square in FIG. The block diagonal component of 2 needs to be zero. For example, the third and fourth perturbation bits cannot be used when the transmission bits arranged by the modulation unit in the third (real part) and fourth (imaginary part) are calculated by the perturbation bit addition unit. It is.

Next, the post-DPC code signal output from the DPC encoding unit 367 is input to the unique signal configuration unit 363. As in the first embodiment, the eigensignal configuration unit 363 inputs the transmission signal obtained by inserting the DRS into the post-DPC code signal to the linear filter multiplication unit 365. The linear filter multiplication unit 365 multiplies the input transmission signal by the linear filter represented by the equation (5-1) or the equation (5-2), and the signal corresponding to each antenna is a power normalization unit. 901 is entered. After calculating the signals addressed to all antennas, the power normalization unit 901 performs power normalization on the signals addressed to all terminals so that the transmission power per one subcarrier and one OFDM symbol is E _tr . The power normalization unit calculates a normalization coefficient g and multiplies the signal corresponding to each antenna. This power normalization does not necessarily need to be performed for each subcarrier and 1 OFDM symbol, and 1 for each subcarrier so that the average transmission power of the plurality of subcarriers is E _tr per 1 subcarrier and 1 OFDM symbol. One g may be calculated, or one g may be calculated with a plurality of subcarriers and a plurality of OFDM symbols so that the transmission power averaged with a plurality of subcarriers and a plurality of OFDM symbols is E _tr .

  The above is the configuration of the base station apparatus. In the present embodiment, it is not necessary to change the configuration of the DPC decoding unit from the first embodiment on the terminal side.

  According to the present embodiment, the present invention can be applied to a VP having high transmission characteristics in MU-MIMO, and Modulo Loss, which has been a factor of characteristic deterioration in VP, can be reduced.

(Sixth embodiment)
In the first to fifth embodiments described above, a constellation such as Gray Mapping that can treat the bits arranged in one modulation symbol as being uncorrelated with each other is used. In the present embodiment, a case of a constellation in which bits arranged in one modulation signal have a correlation will be described.

  In this embodiment, by using a specific constellation, it is possible to obtain better error rate characteristics than when using a constellation in which bits can be handled as uncorrelated. This is due to the incorporation of the BICM-ID method into the present invention.

  This embodiment can be implemented as it is with any configuration of the third to fifth embodiments of the transmission / reception apparatus. However, the operations of the modulation unit of the transmission device and the transmission bit soft estimation unit of the reception device are different from those of the third to fifth embodiments.

  The modulation unit of the transmission apparatus modulates transmission bits using a method in which bits are correlated with each other, such as Set Partitioning illustrated in FIG. 22 instead of Gray Mapping, which is used in the BICM-ID method. Since the operations other than the modulation unit are the same as those in the third to fifth embodiments, the description thereof is omitted.

  Here, a difference between a constellation in which bits are correlated and a constellation in which there is no correlation will be described with reference to FIG. FIG. 26 (a) shows the same constellation as FIG. 22 (a), and FIG. 26 (b) shows a constellation using Gray Mapping. Both are constellations in which two bits are arranged in one modulation signal. In FIG. 26B, if the first bit (the number on the right of the numbers assigned to each bit in FIG. 26) is 0, the signal point is always in the positive position on the real axis, and the first bit is 1 Then, the signal point is always in a negative position on the real axis. In addition, if the second bit (the number on the left in the numbers in FIG. 26) is 0, the signal point is always positive on the imaginary axis, and if the second bit is 1, the signal point is always Negative position on the imaginary axis. Here, the arrangement on each axis does not change depending on whether the first bit or the second bit is 0 or 1.

  On the other hand, in FIG. 26A, when the second bit is 0, if the first bit is 0, the signal point is a positive position on the imaginary axis, but even if the second bit is 0, If the first bit is 1, the signal point is in a negative position on the imaginary axis. In this way, a constellation in which the arrangement of each bit changes depending on whether the other bits are 0 or 1 is referred to as a constellation in which bits are correlated here.

  However, 16QAM and 64QAM using Gray Mapping place multiple bits on the real and imaginary axes, but they can be regarded as being approximately independent, and as a constellation with no correlation between bits. Can be handled.

  This embodiment is an embodiment in which a constellation in which bits are correlated, such as Set Pertitioning, is used instead of a constellation in which bits are not correlated, such as Gray Mapping. The transmission bit soft estimator of the receiving apparatus performs soft estimation of the transmission bits using the following equation instead of equation (3-2).

という因子が追加されている。これは、c番目の送信ビットLLRを算出する場合、c番目以外の全送信ビットの取り得る確率を乗算していることとなる。この因子を追加することによって同じ変調信号に配置されるビット間の相関を考慮した送信ビットの軟推定が可能となる。また、受信装置における繰り返し処理の初回においては、式（３−３）を用いる。 Here, in the formula (6-1), with respect to the formula (3-2),

The factor is added. This means that when the c-th transmission bit LLR is calculated, the probabilities of all transmission bits other than the c-th are multiplied. By adding this factor, it is possible to perform soft estimation of transmission bits considering the correlation between bits arranged in the same modulation signal. In the first iteration process in the receiving apparatus, the expression (3-3) is used.

とする。また、摂動ビット軟推定部は、式（３−４）と同じ式を用いて送信ビットの軟推定を行う。 In addition, since there is no prior information about the transmission bit at the first time,

And Further, the perturbation bit soft estimation unit performs soft estimation of the transmission bit using the same formula as the formula (3-4).

  As described above, the present invention has good compatibility with the BICM-ID scheme, and a BICM-ID scheme that exhibits higher transmission characteristics than Gray Mapping can be introduced simply by changing the operations of the modulation section and the transmission bit soft estimation section.

<Common to all examples>
In the present invention, when calculating various LLRs, it is desirable to calculate LLRs without using approximation as in the above-described formula, but Max-log approximation is performed in consideration of the processing capability of the receiving apparatus, allowable delay time, and the like. The calculation amount in the receiving apparatus may be reduced using approximation such as.

  The program that operates in the mobile station apparatus and the base station apparatus related to the present invention is a program (a program that causes a computer to function) that controls the CPU and the like so as to realize the functions of the above-described embodiments related to the present invention. Information handled by these devices is temporarily stored in the RAM at the time of processing, then stored in various ROMs and HDDs, read out by the CPU, and corrected and written as necessary. As a recording medium for storing the program, a semiconductor medium (for example, ROM, nonvolatile memory card, etc.), an optical recording medium (for example, DVD, MO, MD, CD, BD, etc.), a magnetic recording medium (for example, magnetic tape, Any of a flexible disk etc. may be sufficient. In addition, by executing the loaded program, not only the functions of the above-described embodiment are realized, but also based on the instructions of the program, the processing is performed in cooperation with the operating system or other application programs. The functions of the invention may be realized.

  In the case of distribution in the market, the program can be stored and distributed in a portable recording medium, or transferred to a server computer connected via a network such as the Internet. In this case, the storage device of the server computer is also included in the present invention. Moreover, you may implement | achieve part or all of the mobile station apparatus and base station apparatus in embodiment mentioned above as LSI which is typically an integrated circuit. Each functional block of the mobile station apparatus and the base station apparatus may be individually made into a processor, or a part or all of them may be integrated into a processor. Further, the method of circuit integration is not limited to LSI, and may be realized by a dedicated circuit or a general-purpose processor. In addition, when an integrated circuit technology that replaces LSI appears due to progress in semiconductor technology, an integrated circuit based on the technology can also be used.

  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 not departing from the gist of the present invention are also claimed. Included in the range.

  The present invention is applicable to mobile communication devices.

DESCRIPTION OF SYMBOLS A ... Transmission side apparatus, 1 ... DPC encoding part, 3 ... Code | symbol part, 5 ... Interleaving part, 7 ... Perturbation bit addition part, 11 ... Generation matrix memory | storage part, 15 ... Transmission signal generation part, 17 ... Modulation part, 19 ... Interference calculation unit, 21 ... dither sequence addition unit, 23 ... perturbation vector addition unit, 25 ... perturbation bit calculation unit, 31 ... DPC decoding unit, 33 ... dither sequence subtraction unit, 35 ... transmission bit soft estimation unit, 37 ... perturbation Bit soft estimator, 41 ... Check Node, 43 ... Check matrix storage, 45 ... Deinterleave, 47 ... Interleave, 51 ... Subtractor, 53 ... Decode, 55 ... Hard decision, B ... Base station, 101a · b ... antenna, 103 ... base station transmitter, 105 ... base station receiver, 111a · b ... radio transmitter, 113a · b ... GI inserter, 115a · b ... IFFT unit, 117a · b ... frame configuration unit , 21 ... CRS generator, 131a.b ... radio receiver, 133a.b ... GI remover, 135a.b ... FFT unit, 137 ... propagation state information acquisition unit, 151 ... linear filter calculator, 153 ... encoder, 155: Modulation unit, 157 ... Interference calculation unit, 161 ... DRS generation unit, 163a ... Eigen signal configuration unit, 163b ... Eigen signal configuration unit, 165 ... Linear filter multiplication unit, 167 ... DPC encoding unit, 201 ... Antenna, 203 ... GI removal unit, 207 ... FFT unit, 211 ... signal separation unit, 215 ... propagation path estimation unit, 217 ... propagation path state information generation unit, 221 ... IFFT unit, 223 ... GI insertion unit, 225 ... wireless transmission unit, 227 ... Propagation path compensator, 231 ... demodulator, 233 ... decoder, 235 ... DPC decoder, 411 ... real part / imaginary part separator, 415 ... real part / imaginary part combiner, 451 ... DPC decoder , 453 ... real part / imaginary part separation part, 603 ... coefficient multiplication part, 653 ... coefficient multiplication part, 701, 801 ... coefficient calculation part, 801 ... coefficient, 901 ... power normalization part, 903 ... perturbation vector calculation part.

Claims (17)

A transmitting device in a communication system in which a transmitting device transmits a signal to at least one receiving device,
A code unit that generates a code bit by error correction coding, a modulation unit that generates a modulated signal by modulating a transmission bit, and
A perturbation vector addition unit that generates a transmission signal by adding a perturbation vector that is a signal that is an integral multiple of a signal of a known magnitude in the reception device and the transmission device to the modulation signal;
A perturbation bit calculator for calculating a perturbation bit based on the perturbation vector;
A perturbation bit addition unit for calculating a transmission bit by adding the perturbation bit to at least a part of the sign bit;
A transmission device comprising: The perturbation bit calculation unit includes:
Calculating a perturbation bit based on a value obtained by dividing the perturbation vector by the signal of the known magnitude;
The transmission device according to claim 1. The perturbation bit calculation unit includes:
Calculating a perturbation bit based on whether the divided value is odd or even;
The transmission device according to claim 2. An interference subtraction unit that subtracts an interference signal received by the reception signal from the modulation signal;
The transmission device according to any one of claims 1 to 3. A coefficient multiplier that multiplies the interference signal received by the received signal from the modulated signal by a positive coefficient less than 1, and an interference subtractor that subtracts the signal multiplied by the coefficient;
The transmission device according to any one of claims 1 to 3. The perturbation vector addition unit
Generating the transmission signal by Modulo operation;
The transmission device according to claim 4, wherein: The transmitting device transmits transmission signals addressed to a plurality of receiving devices simultaneously at the same frequency,
A linear filter multiplier that multiplies the transmission signals calculated for the plurality of receiving devices by a linear filter;
The transmission device according to any one of claims 1 to 3. Having a perturbation vector calculation unit for calculating the perturbation vector to be added to the receiving device based on the linear filter and the modulation signal transmitted to the receiving device;
The transmission apparatus according to claim 7. A receiving device in a communication system in which a transmitting device transmits a signal to at least one receiving device,
A perturbation bit soft estimator for estimating a perturbation bit and calculating a perturbation bit soft estimate based on a received signal;
A transmission bit soft estimator that estimates a transmission bit and calculates a transmission bit soft estimate based on a received signal;
Check Node for calculating a sign bit soft estimate based on the perturbation bit soft estimate and the transmission bit soft estimate;
A decoding unit that performs decoding based on the code bit soft estimate;
A receiving apparatus comprising: The decoding unit further calculates a code bit soft estimate,
The CheckNode further includes
A transmission bit soft estimation value is calculated based on the code bit soft estimation value and the perturbation bit soft estimation value, and based on the code bit soft estimation value, the perturbation bit soft estimation value, and the transmission bit soft estimation value, a new Calculating the perturbation bit soft estimate,
The receiving device according to claim 9. The transmission bit soft estimation unit regards a point obtained by translating a received signal in the real part or imaginary part direction by an integral multiple of a signal having a known size in the receiving apparatus and the transmitting apparatus as the same received signal point. Calculate the transmitted bit soft estimate,
The perturbation bit soft estimation unit calculates a perturbation bit soft estimation value by regarding a point obtained by translating a reception signal in the real part or imaginary part direction by an even multiple of the known size as the same reception signal point. To do,
The receiving device according to claim 9 or 10. For the received signal, the decoding unit
Having a coefficient multiplier that multiplies positive coefficients less than 1;
The receiving device according to claim 9, wherein the receiving device is characterized in that: A communication system in which a transmission device transmits a signal to at least one reception device, a code unit that generates a code bit by error correction coding, a modulation unit that modulates the transmission bit and generates a modulation signal;
A perturbation vector addition unit that generates a transmission signal by adding a perturbation vector that is a signal that is an integral multiple of a signal of a known magnitude in the reception device and the transmission device to the modulation signal;
A perturbation bit calculator for calculating a perturbation bit based on the perturbation vector;
A perturbation bit addition unit for calculating a transmission bit by adding the perturbation bit to at least a part of the sign bit;
A transmission device comprising:
A perturbation bit soft estimator for estimating a perturbation bit and calculating a perturbation bit soft estimate based on a received signal;
A transmission bit soft estimator that estimates a transmission bit and calculates a transmission bit soft estimate based on a received signal;
Check Node for calculating a sign bit soft estimate based on the perturbation bit soft estimate and the transmission bit soft estimate;
A decoding unit that performs decoding based on the code bit soft estimate;
A receiving device comprising:
A communication system comprising: A transmission method in a communication system in which a transmission device transmits a signal to at least one reception device,
Generating a code bit by error correction coding; modulating a transmission bit to generate a modulated signal; and
Adding a perturbation vector, which is a signal that is an integral multiple of a signal of a known magnitude in the receiving apparatus and the transmitting apparatus, to the modulation signal to generate a transmission signal;
Calculating a perturbation bit based on the perturbation vector;
Adding the perturbation bit to at least a part of the sign bit to calculate a transmission bit;
Including sending method. A receiving method in a communication system in which a transmitting device transmits a signal to at least one receiving device,
Calculating a perturbation bit soft estimate by estimating a perturbation bit based on the received signal;
Calculating a transmission bit soft estimate by estimating a transmission bit based on a received signal;
Calculating a sign bit soft estimate based on the perturbation bit soft estimate and the transmission bit soft estimate;
Decoding based on the code bit soft estimate;
Including receiving method. A processor used in a transmission apparatus in a communication system in which the transmission apparatus transmits a signal to at least one reception apparatus,
A code unit that generates a code bit by error correction coding, a modulation unit that generates a modulated signal by modulating a transmission bit, and
A perturbation vector addition unit that generates a transmission signal by adding a perturbation vector that is a signal that is an integral multiple of a signal of a known magnitude in the reception device and the transmission device to the modulation signal;
A perturbation bit calculator for calculating a perturbation bit based on the perturbation vector;
A perturbation bit addition unit for calculating a transmission bit by adding the perturbation bit to at least a part of the sign bit;
A processor comprising: A processor used in a receiving apparatus in a communication system in which a transmitting apparatus transmits a signal to at least one receiving apparatus,
A perturbation bit soft estimator for estimating a perturbation bit and calculating a perturbation bit soft estimate based on a received signal;
A transmission bit soft estimator that estimates a transmission bit and calculates a transmission bit soft estimate based on a received signal;
Check Node for calculating a sign bit soft estimate based on the perturbation bit soft estimate and the transmission bit soft estimate;
A decoding unit that performs decoding based on the code bit soft estimate;
A processor comprising:

Images