JP2010028384A - Radio transmitting method and apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a radio transmitting apparatus which alleviates increase of dynamic ranges of receiving signals in a multiuser radio receiving apparatus while suppressing increase of transmission power to allow reception by an inexpensive receiving apparatus with a small dynamic range. <P>SOLUTION: The radio transmitting apparatus includes: a feedback processing part which performs feedback processing for subtracting a signal fed back by multiplying a gain from a signal to be transmitted; an operation part which performs an operation for reducing the increase of the transmission power by the feedback processing to the signal after the feedback processing; a beam former which performs beam forming processing to the signal after the operation; a transmitting part which transmits transmission signals including the signal after the beam forming processing; a computation part which computes an evaluation value expressing the dynamic range of a received signal in a radio receiving apparatus which receives the transmission signals; and an assignment part which assigns the signal to be transmitted to at least one transmission stream so that the dynamic range of the received signal is settled within the minimum or a reception dynamic range of the radio receiving apparatus based on the evaluation value. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention particularly relates to a wireless transmission method and apparatus for transmitting to a plurality of wireless terminals.

  A Spatial Division Multiple Access (SDMA) system that multiplexes a plurality of users at the same frequency and the same time using a plurality of transmission antennas is known as a technique that can effectively use the frequency. Applying this technology to a wireless communication system that performs communication between a base station and a plurality of users (wireless terminals), allowing a plurality of users to receive only transmission signals for each user without interfering with each other It has been known.

  The channel inversion (CI) scheme is a standard for reducing interference to zero, and is an inverse matrix of a matrix (hereinafter referred to as a channel matrix) representing a plurality of propagation path states between a plurality of transmission antennas and a plurality of users in a base station, for example. Interference suppression is performed by performing beam forming by multiplying the transmission signal by a weight W corresponding to. This standard is called channel inversion (CI) because SDMA is realized using an inverse matrix of the channel matrix.

  However, if the inverse matrix of the channel matrix is used, the transmission signal level increases in some cases, and the transmission power must be lowered to match the rated transmission power. As described above, the CI method has a problem in that good characteristics cannot be obtained by performing transmission with lower transmission power.

Non-Patent Document 1 (G. Caire and S. Shamai (Shitz), “On the achievable throughput of a multiantenna Gaussian broadcast channel,” IEEE Trans. On Info. Theory, vol. 49, no. 7, pp. 1691-1706 , Jul. 2003.) proposes ZF (zero-forcing) -DPC (Dirty Paper Coding) which solves the problem of the CI method. In the ZF-DPC system, Q obtained by QR-decomposing ^HH , which is Hermitian transpose of the channel matrix H, is used as a weight W for beamforming. Since Q is an orthogonal matrix, there is no increase in transmission power, which is a problem in the CI method. When the transmission signal is multiplied by the weight W by beamforming, the wireless terminal of user 1 receives only the transmission signal for user 1, and the wireless terminal of user 2 receives both the transmission signal for user 1 and the transmission signal for user 2. Will be received.

  In this case, a signal is transmitted to the user 1 using a transmission beam that can obtain the maximum gain in the wireless terminal of the user 1. Therefore, since it is not necessary to suppress interference for the user 1, it is possible to transmit with the beam with the maximum ratio combining, and the characteristics are improved as compared with the CI method. At this time, the diversity order of the wireless terminal of user 1 is 1. Here, since the transmission beam for the user 1 does not consider the wireless terminal of the user 2, the transmission signal for the user 1 reaches the user 2 as well. On the other hand, the transmission signal for the user 2 needs to suppress the interference to the wireless terminal of the user 1 and therefore does not reach the wireless terminal of the user 1 but does not reach the wireless terminal of the user 2 with the maximum gain. The diversity order of the wireless terminal of user 2 is 1.

  By the way, since the wireless terminal of the user 2 receives interference from the transmission signal for the user 1 in this state, the ZF-DPC performs precoding by feedback processing on the transmission signal. Thus, by using both beamforming and precoding, the wireless terminals of user 1 and user 2 can communicate with the base station without interfering with each other. In this case, the diversity order of the wireless terminal of the user 1 is 2, and the diversity order of the wireless terminal of the user 2 is 1. On the other hand, in the CI system, the diversity order is 1 for all users. Therefore, it can be said that ZF-DPC is a method superior to CI. However, ZF-DPC has a problem that transmission power increases when precoding is performed only for interference cancellation.

In contrast, Non-Patent Document 2 (C. Windpassinger, R. Fischer, T. Vencel, and JB Huber, “Precoding in multiantenna and multiuser communications,” IEEE Trans. Wireless Communication, vol. 3, no. 4, pp. 1305-1316, July 2004.) proposes a transmission power reduction method using Tomlinson-Harashima Precoding (THP). This is a method for reducing transmission power by performing a modulo operation (referred to as modulo reduction in Non-Patent Document 2) as shown by THP on a transmission signal composed of complex numbers. . Hereinafter, this method is referred to as multi-user THP. Multi-user THP can improve reception quality over a general SDMA system while multiplexing users at the same time and frequency without increasing transmission power.
G. Caire and S. Shamai (Shitz), “On the achievable throughput of a multiantenna Gaussian broadcast channel,” IEEE Trans. On Info. Theory ,. vol. 49, no. 7, pp. 1691-1706, Jul. 2003 . C. Windpassinger, R. Fischer, T. Vencel, and JB Huber, “Precoding in multiantenna and multiuser communications,” IEEE Trans. Wireless Communication, vol.3, no.4, pp.1305-1316, July 2004.

  By the way, in the multi-user THP disclosed in Non-Patent Document 2, the received signal changes depending on the signal component added by the modulo operation. That is, depending on the channel state, the received signal may have a large value or a small value depending on the added signal component. As described above, depending on the case, the reception power fluctuates greatly, so that the dynamic range (reception dynamic range) of the reception signal that can be received by the reception apparatus becomes very large, and reception by an inexpensive reception apparatus is impossible. There is a problem that it becomes.

  The present invention suppresses an increase in transmission power and reduces an increase in the dynamic range of a received signal in a multi-user radio reception apparatus, thereby enabling radio transmission with an inexpensive radio reception apparatus having a small reception dynamic range. It is an object to provide a method and apparatus.

  According to one aspect of the present invention, a method of transmitting a transmission signal to a plurality of wireless reception devices at the same frequency and at the same time using a plurality of antennas, the transmission signal is reduced in order to reduce an increase in transmission power. Adding a perturbation vector so as to be within a predetermined level, and a dynamic range of the receiver that receives the transmission signal is minimized, or a dynamic range of the reception signal is a reception dynamic range of the radio reception device Assigning the signal to be transmitted to at least one transmission stream so as to be contained within a wireless transmission method.

  According to another aspect of the present invention, performing a feedback process of multiplying a signal to be transmitted by a gain, feeding back a signal that causes interference in a wireless reception device, and subtracting the signal that causes interference from the signal to be transmitted; Performing a calculation for reducing an increase in transmission power due to the feedback process on the signal after the feedback process, and a beam for reducing interference remaining in the feedback unit for the signal after the calculation Performing a forming process; transmitting a transmission signal including the signal after the beam forming process; calculating an evaluation value representing a dynamic range of a reception signal in a wireless reception device that receives the transmission signal; Based on the evaluation value, the dynamic range of the received signal is minimized. Or allocating the signal to be transmitted to at least one transmission stream so that the dynamic range of the received signal falls within the reception dynamic range of the radio reception apparatus. Is provided.

  According to still another aspect of the present invention, feedback processing for performing feedback processing for multiplying a signal to be transmitted by a gain, feeding back a signal that causes interference in a radio reception apparatus, and subtracting the signal that causes interference from the signal to be transmitted A calculation unit that performs an operation to reduce an increase in transmission power due to the feedback process on the signal after the feedback process, and reduces an interference remaining in the feedback unit with respect to the signal after the calculation An evaluation value representing a dynamic range of a received signal in a beamformer for performing beamforming processing, a transmission unit for transmitting a transmission signal including the signal after the beamforming processing, and a wireless reception device for receiving the transmission signal A calculation unit for calculating, and a dynamic level of the received signal based on the evaluation value; An allocation unit that allocates the signal to be transmitted to at least one transmission stream so that the dynamic range of the reception signal is within the reception dynamic range of the wireless reception device. Is provided.

  According to the present invention, an increase in the dynamic range of a received signal in a multiuser radio reception apparatus can be reduced while suppressing an increase in transmission power, and reception by an inexpensive radio reception apparatus having a small reception dynamic range becomes possible.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
(First embodiment)
First, schematic configurations of the wireless transmission device and the wireless reception device according to the first embodiment will be described.
<Wireless transmitter>
FIG. 1 shows a wireless transmission device using multi-user THP according to the first embodiment. In this embodiment, an example in which three users (wireless terminals) are assigned to one transmission stream using three transmission antennas will be described. It is possible to further increase the number of transmission antennas and assign each user to two or more transmission streams, or it is possible to increase the number of users. Here, the transmission stream is a vector element of a transmission signal, which will be described later in <Detailed description>. In other words, the “assignment to transmission stream” process is a process of assigning which number of user signals to which element of the transmission signal vector.

  The weight sequence calculation unit 100 calculates a weight R for the feedback unit 105 and a weight Q for the beamformer 106 using channel information (propagation path information) notified by feedback from a receiving device (not shown). The calculation results of the weights R and Q are also sent to the scheduling unit 101.

  On the other hand, data signals D <b> 1 to D <b> 3 to be transmitted to users 1 to 3 are input to the user assignment unit 102. The user allocation unit 102 is controlled by the scheduling unit 101 and performs a process of allocating data signals D1 to D3 to be transmitted to the users 1 to 3 to any one of the transmission streams. The data signals D1 to D3 are obtained by various digital modulation schemes known by a modulation unit (not shown), such as BPSK (Binary Phase Shift Keying), QPSK (Quadrature Phase Shift Keying), ASK (Amplitude Shift Keying), and FSK (Frequency). It is assumed that modulation is performed by a modulation scheme such as Shift Keying), 16 QAM (16 Quadrature Amplitude Modulation), or 64 QAM.

  The transmission signal s output from the user assignment unit 102 is processed by a feedback processing / modulo calculation unit surrounded by a broken line, which includes a subtraction unit 103, a modulo calculation unit 104, and a feedback unit 105. That is, the transmission signal s is subjected to feedback processing by the subtraction unit 103 and the feedback unit 105. The feedback unit 105 performs a process of multiplying the signal output from the subtraction unit 103 by a certain gain. Further, the modulo operation unit 104 performs modulo operation on the signal after the feedback processing. The modulo operation here is slightly different from a general (narrowly) modulo operation, and the details will be described later. The data signal v after the modulo operation unit is input to the beam former 106.

  In the beam former 106, processing for forming transmission beams for a plurality of users, that is, beam forming processing is performed on the data signal v after the modulo calculation. When using Orthogonal Frequency-Division Multiplexing (OFDM) transmission or Orthogonal Frequency Division Multiplexing Access (OFDMA), the data signal after beamforming is processed by inverse fast Fourier transformer (IFFT). After being converted into an OFDM signal by the unit 107, it is converted into an analog signal by a digital-analog converter (DAC) 108 and input to a plurality (n) of transmission RF (Radio Frequency) units 109-1 to 109 -n. .

  In the case of single carrier transmission, that is, when multicarrier transmission such as OFDM transmission or OFDMA transmission is not used, the IFFT unit 107 is not necessary, and the output signal from the beam former 106 is directly input to the DAC 108. In either case, a digital filter for band limitation may be used before the DAC 108.

  The transmission RF units 109-1 to 109-n include a frequency converter (upconverter), a power amplifier, and a filter as necessary. The pilot signal and the data signal converted into the analog signal from the DAC 108 are up-converted to a frequency in the RF band in the transmission RF units 109-1 to 109-n, and further subjected to power amplification, and then transmitted to the transmission antennas 110-1 to 110-1. 110-n.

<Wireless receiver>
FIG. 2 shows a wireless reception device according to the first embodiment. The radio reception apparatus shown in FIG. 2 corresponds to the radio transmission apparatus shown in FIG. The RF signal transmitted from the wireless transmission device in FIG. 1 is received by the reception antenna 301. A reception signal 31 output from the reception antenna 301 is input to the reception RF unit 302.

  The reception RF unit 302 includes a low noise amplifier (LNA), a frequency converter (down converter), and a filter as necessary. The reception signal 31 is amplified by the reception RF unit 302 and then down-converted from the RF band to, for example, a baseband frequency.

  A baseband received signal output from the reception RF unit 302 is converted from an analog signal to a digital signal by an analog-digital converter (ADC) 303, and then a fast Fourier transformer (FFT) unit 304. Is demodulated (subcarrier demodulation). An output signal from the FFT unit 304 is input to the modulo arithmetic unit 305 and the channel estimation unit 306. Output signals from the modulo arithmetic unit 305 and the channel estimation unit 306 are input to the demodulation unit 307.

  Here, in response to the fact that the wireless transmission device of FIG. 1 uses multicarrier transmission such as OFDM transmission, an FFT unit 304 is provided in the wireless reception device, but in the case of single carrier transmission, FFT is performed. The unit 304 is not necessary, and an output signal from the ADC 303 is directly input to the modulo arithmetic unit 305 and the channel estimation unit 306. In either case, a digital filter for band limitation may be used after the ADC 303.

  The modulo operation unit 305 performs modulo operation corresponding to the modulo operation unit 104 in FIG. Channel estimation section 306 performs channel estimation between transmission antennas 110-1 to 110-n of the wireless transmission device and each wireless reception device, that is, estimation of a propagation path response (hereinafter referred to as channel response). The demodulation unit 307 performs channel equalization on the output signal from the FFT unit 304 based on the channel estimation value from the channel estimation unit 306, and then performs demodulation corresponding to the modulation of the modulation units 101 and 102 in FIG. Processing, that is, processing for demapping the modulated signal to the signal point is performed.

  The channel estimation unit 306 includes, for example, two memories 401 and 402, an orthogonal processing unit 403, and a division unit (bit shift unit) 404 as shown in FIG. Detailed processing of the channel estimation unit 306 will be described later.

<Detailed explanation>
Next, the wireless transmission device and the wireless reception device according to the present embodiment will be described in detail together with details of conventional CI, ZF-DPC, and multiuser THP.
The radio transmission apparatus according to the present embodiment is provided in a base station of a radio communication system having, for example, a base station and radio terminals possessed by a plurality of users. A wireless communication system uses an SDMA scheme in which a plurality of users (wireless terminals) are multiplexed at the same frequency and the same time using a plurality of antennas.

The base station has two transmitting antennas Tx1 and Tx2 (for example, corresponding to antennas 110-1 and 110-2 in FIG. 1), and the wireless terminals of user 1 and user 2 each have one receiving antenna Rx1 and Rx2. Suppose you have it. The transmission signals to the wireless terminals of user 1 and user 2 respectively

And the noise signal received by the wireless terminals of user 1 and user 2

Then, the signal y received by the wireless terminals of user 1 and user 2 is

Can be written. Here, H is a channel matrix between the base station and the wireless terminal, h ₁₁ is a channel response from the transmitting antenna Tx1 to the receiving antenna Rx1 of the wireless terminal of the user 1, and h ₁₂ is a channel from the transmitting antenna Tx2 to the receiving antenna Rx1. Response, h ₂₁ indicates a channel response from the transmission antenna Tx1 to the reception antenna Rx2 of the wireless terminal of the user 2, and h ₂₂ indicates a channel response from the transmission antenna Tx2 to the reception antenna Rx2.

Here, for the transmission signal s using a beamformer

Multiply the weight W

Thus, since the wireless terminals of user 1 and user 2 can receive only s ₁ and s ₂ without interference with each other, SDMA can be realized. This is the CI method described above.

On the other hand, ZF-DPC described in Non-Patent Document 1 will be described with reference to FIGS. 4 and 5. In ZF-DPC, as described above, Q obtained by QR decomposition of Hermitian transpose H ^H of channel matrix H is used. In particular,

Is used as the weight W. Here, since Q is an orthogonal matrix and a unitary matrix, an increase in transmission power that causes a problem in the CI method does not occur.

When the transmission signal s is multiplied by the weight W by beam forming, the reception signals of the wireless terminals of the user 1 and the user 2 are as follows.

Therefore, the wireless terminal of the user 1 receives only the transmission signal s ₁ for the user 1, the wireless terminal of the user 2 receives both the transmission signal s ₂ of the transmission signal s ₁ and the user 2 for for user 1.

FIG. 4 shows an image of the beam created by the weight W at this time. In the example of FIG. 4, the transmission signal s ₁ for the user 1 is transmitted with a beam that can obtain the maximum gain in the wireless terminal of the user 1. Since it is not necessary to consider interference suppression for the beam for user 1, it can be a beam for maximum ratio combining diversity, and has better characteristics than the CI method.

The reception signal of the wireless terminal of user 1 is y ₁ = r ₁₁ ^* s ₁ + n ₁ . At this time, the diversity order of the wireless terminal of the user 1 is 2. Since the beam for the user 1 does not consider the wireless terminal of the user 2, the transmission signal s1 for the user 1 reaches the wireless terminal of the user 2 as well. On the other hand, transmission signal s ₂ of the user 2 for, it is necessary to suppress the interference to the user 1 of the wireless terminal. As a result, the transmitted signal s ₂ of the user 2 for, although not reach the wireless terminal of the user 1, it does not reach the maximum gain in the wireless terminal of the user 2, the diversity order of the user 2 the wireless terminal becomes 1.

By the way, the wireless terminal of the user 2 will receive interference from the transmission signal s ₁ for the user 1 in this state. That is, the received signal of the wireless terminal of user 2 is y ₂ = r ₂₁ ^* s ₁ + r ₂₂ ^* s ₂ + n ₂ , where r ₂₁ ^* s ₁ is an interference component from s ₁ . Therefore, in ZF-DPC, as shown in FIG. 5, transmission signals s ₁ and s ₂ are subjected to precoding including feedback processing, and then beam forming is performed. In FIG. 5, a precoding signal s ′ represented by the following equation is generated.

When such precoding (feedback processing) is performed, a final transmission signal x ′ obtained by performing beamforming with a weight W on the precoding signal s ′ is expressed by the following equation.

The received signal y ′ at this time can be written as follows.

Where B is

  From equation (10), in ZF-DPC, user 1 and user 2 can communicate without mutual interference by using both beamforming and precoding.

That is, as shown in FIG. 5, the received signal of the wireless terminal of user 1 is y ₁ = r ₁₁ ^* s ₁ + n ₁ as in the case of FIG. 4 where precoding is not performed. On the other hand, the received signal of the wireless terminal of user 2 is y ₂ = r ₂₂ ^* s ₂ + n ₂ , and the interference component r ₂₁ ^* s _{1 is cancelled.} Thus, by using both beamforming and precoding, the wireless terminals of user 1 and user 2 can communicate with the base station without mutual interference.

  By the way, when precoding, that is, feedback processing is performed on the transmission signal as shown in Equation (10), the signal s ′ after the feedback processing is as shown in Equation (8).

Here, even when the powers of s ₁ and s ₂ are normalized as 1, s ₂ −s ₁ r ₁₂ ^* / r ₂₂ ^* (s ₂ −s ₁ r ₁₂ ^{(1) *} / r ⁾ in Expression (8) ₂₂ ^{(1) *} ) is likely to be greater than 1, which increases the transmission power. In order to avoid this problem, as described above, Non-Patent Document 2 proposes a transmission power reduction method using THP, that is, a method of performing a modulo operation on a transmission signal composed of complex numbers. Hereinafter, the principle and problems of the transmission power reduction method proposed in Non-Patent Document 2 will be described with reference to FIGS. FIG. 6 shows a constellation of transmission signal points on the IQ plane, and FIG. 7 shows a constellation of reception signal points on the IQ plane.

First, the amplitude A of s ₁ and s ₂ of the transmission signal before canceling is expressed as follows.

Here, M is the modulation multi-value number, m is a coefficient for normalizing the power to 1, and in the case of QPSK, M = 4 and m = √2. Therefore, the signal point of QPSK is

It becomes. Now, when transmitting the upper left signal point s2 = (− 1 / √2 + j / √2) in FIG. 6, it is assumed that the signal point s2 has moved to the upper right white point s2 ′ by precoding. At this time, in the multi-user THP, a real number N = 2√M / m = 2√2 is an integer multiple of the precoded signal s ₂ -s ₁ r ₁₂ ^{(1) *} / r ₂₂ ^{(1) *} . The power is reduced by subtracting or adding on the imaginary axis and the axis so that the amplitude is within ± √M / m = ± 2√2 for both the I phase and the Q phase. Since this process is similar to modulo arithmetic, C. Windpassinger and colleagues call it modulo reduction.

Now, subtract N once for the I phase and N once for the Q phase in the modulo operation,

Suppose that the amplitude ± √M / m becomes ± √2 or less. That is, the point s ₂ 'is to moved to s ₂ "amplitude ± √M / m = ± 2√2 within point as shown by an arrow. In this case, by performing a modulo operation (modulo reduction), transmission symbol Falls within ± √M / m = ± 2√2, so transmission can be performed without increasing the transmission power.

On the other hand, if the noise component is ignored on the receiving side,

Is received. Therefore, when the signal of Expression (15) is divided by the gain r ₂₂ ^* corresponding to the channel response, the received signal becomes s ₂ −N−jN, which is a white dot in FIG. Since this received signal exceeds the amplitude ± √M / m = ± 2√2 before the feedback processing, the signal on the receiving side is obtained by adding N in Equation (13) once for the I phase and Q phase. it becomes possible to recover the s _2. This method is the multi-user THP described above, and has the characteristics that it is possible to multiplex users at the same time and the same frequency without increasing the transmission power and to improve the reception quality over the conventional SDMA method.

  Here, as shown in the equation (15), the received signal changes depending on the signal component (−N−jN) added by the modulo operation. That is, depending on the state of the channel (propagation path), the value of equation (15) may increase or decrease. In this way, the received power fluctuates greatly depending on the situation, so that the dynamic range of the received signal input to the receiving device becomes very large, and reception by an inexpensive receiving device becomes impossible. There is.

In the present embodiment, such an increase in the dynamic range is reduced as follows. Hereinafter, the principle of reducing the increase in the dynamic range of the received signal in this embodiment will be described. Here, a case where transmission is performed by assigning three users to one transmission stream is considered as an example. Considering even the modulo operation, the received signal in the multi-user THP can be expressed as the following equation when expressed as a vector.

  Here, d is a vector of signal components added by modulo calculation corresponding to −N−jN in the equation (14), and is added so that x falls within ± √M / m. Hereinafter, this vector d will be referred to as a perturbation vector.

  In the equation (16), s1 to s3 respectively represent transmission signals to the users 1 to 3, and the first transmission stream to the third transmission stream in order from the top element of the transmission signal vector s of the equation (16). I will call it. That is, Expression (16) is a case where the transmission signal of each user is assigned to each transmission stream so that the user number and the transmission stream number are the same. According to Equation (16), it can be seen that the dynamic range of the received signal depends on the magnitude of the perturbation vector d (d1, d2, d3). Next, how the perturbation vector d is added on the transmission side will be described.

As shown in FIG. 1, when the output of the feedback processing / modulo operation unit (the signal after the modulo operation) is set as v = B ⁻¹ (s + d), each element vi of v is calculated in order from v1. It can be written as follows:

Here, the processing of the modulo operation referred to in the present embodiment is defined again as follows.

MOD () ^{def in} Expression (18) represents a modulo operation in a narrow sense. Equation (18) shows that the perturbation vector d _i is added to the signal si ′ after the feedback processing so that the signal vi after the modulo calculation is within ± √M / m for both the I phase and the Q phase. That means. Moreover, the process of Formula (18) is independently performed in I phase and Q phase.

  In the first transmission stream, that is, the transmission signal to the user 1 calculated in the first row of Expression (17), since there is no signal component to be canceled, the perturbation vector is d1 = 0, and the signal after the modulo calculation is v1 = s1'-s1. The transmission signal to the user 2 calculated in the second transmission stream, that is, the second row of the equation (17) is the interference component v1 (r * of the transmission signal to the user 1 with respect to the transmission signal s2 to the user 2. The perturbation vector d2 is added so that s1 ′ minus 12 / r * 22) is within ± √M / m, and v2 is generated. Therefore, for example, the real component of d2 is real (d2) = real (s'2) -real (v2) from equation (18). Note that real () represents a real number component. Similarly, the real component of the third transmission stream, that is, the perturbation vector d3 of the transmission signal to the user 3 calculated in the third row of Expression (17) is real (d3) = real (s'3) -real ( v3).

  FIG. 8 shows the relationship between the signal si ′ after the feedback processing and the perturbation vector di. As shown in FIG. 18, it can be seen that the perturbation vector di changes stepwise with respect to the signal si ′ after the feedback processing due to the characteristics of the modulo operation.

Here, in the case of the signal stream s3 to the user 3, the real part of the perturbation vector d3 is the largest.

You can think of when is the largest. From FIG. 8, it can be considered that the maximum value of di is obtained when the maximum value of si ′, because di increases with the increase of si ′. Therefore, the maximum value of the real part of the perturbation vector d3 is as follows.

In general, the maximum value of the real component of the perturbation vector dk in the k-th transmission stream can be written as:

That is, in general, a signal received by a user's wireless terminal assigned to the i-th transmission stream is:

It becomes. Note that the second term and the third term on the right side of Equation (22) represent the maximum values of the real component and the imaginary component of the perturbation vector di, respectively. Therefore,

It can be seen that if the maximum value di, max of the perturbation vector di is small, the dynamic range of the received signal is small, and if di, max is large, the dynamic range of the received signal is large. Therefore, it can be seen that di, max in equation (23) should be reduced in order to reduce the dynamic range of the received signal.

By the way, di, max is

Affected by. This is a value calculated from the channel matrix based on Equation (6). In Expression (16), User 1 is assigned to the first transmission stream, User 2 is assigned to the second transmission stream, and User 3 is assigned to the third transmission stream. However, if the user assignment is changed due to the nature of Expression (6), Expression (24) ) Value also changes. Therefore, the value of di, max when each user is assigned to each transmission stream is calculated in advance, and each user may be assigned to each transmission stream so that di, max is minimized. In the present embodiment, assigning each user to each transmission stream so that di and max are minimized is called scheduling. Scheduling is performed by the scheduling unit 101, and actual allocation is performed by the user allocation unit 102. The above is the principle of reducing the dynamic range of the received signal in this embodiment.

  Next, details of each block in FIG. 1 will be described. The weight calculation unit 100 and the scheduling unit 101 calculate the user allocation that minimizes the dynamic range of the received signal while sharing information with each other (in other words, the di, max is reduced), and based on this The user allocation unit 102 allocates each user to each transmission stream. Hereinafter, detailed operations of the weight calculation unit 100 and the scheduling unit 101 will be described with reference to the flowchart shown in FIG.

  First, the number of user allocation candidates, that is, the number of combinations of a plurality of users and a plurality of transmission streams is determined (step S1). When three users are assigned to three transmission streams, the number of combinations is 3 × 2 × 1 = 6. Subsequently, six loops corresponding to the number of combinations start (step S2).

  When the loop starts, first, a row of the channel matrix H is assigned to each user according to the combination of the user and the transmission stream. For example, in the initial state, allocation is performed such that the channel of user 1 is in the first row, the channel of user 2 is in the second row, and the channel of user 3 is in the third column (step S3).

Channel assignment is calculated using a matrix such as: For example, the apparent propagation path H ′ after the assignment is made can be written as:

Here, when assigning user numbers and stream numbers in reverse order, such as assigning user 3 to stream 1, user 2 to stream 2, and user 1 to stream 3, P is given by the following equation.

Here, P is a matrix with one “1” and “0” in other rows and columns. Scheduling, ie, stream allocation, is how to set this P. Next, a weight is calculated by the weight calculation unit 100 using the apparent channel matrix H ′. Specifically, R ′ is calculated based on Expression (6), and B ′ is calculated based on Expression (11) (Step S4).

  Next, the scheduling unit 101 calculates an evaluation value representing the dynamic range of the received signal (step S5). That is, for example, the maximum value di, max of the perturbation vector shown in Expression (23) is calculated. For the evaluation value, the absolute value of Equation (23) may be used, or the power may be calculated. For example, when there are three transmission streams, di and max corresponding to the third transmission stream (final transmission stream) for canceling the most symbols may be used as an evaluation value, or all transmission streams are supported. The maximum di, max may be used as the evaluation value. Also, the final evaluation value may be a value obtained by weighting and combining the di and max values corresponding to each transmission stream. Alternatively, only the equation (24) that changes depending on the assignment in the equation (23) may be calculated and compared.

  Next, the user assignment unit 102 compares the value of the evaluation value calculated in step S5 with a threshold value (step S6). As the threshold value, an evaluation value in a combination calculated in the past is used. Since di, max should be smaller, if the evaluation value (threshold value) in the previously calculated combination is better than the evaluation value in the currently calculated combination, the process returns to step S3 and proceeds to the next combination. .

  If the evaluation value in the combination currently calculated is better than the evaluation value (threshold value) in the combination calculated in the past, the current evaluation value is stored (step S7), and the current combination is used. The configured channel matrix and weight are also stored (step S8).

  If the processing of steps S3 to S8 is not completed for all combinations (sets) of the user and the transmission stream, the processing of S3 to S8 is performed for the remaining sets, and the processing is completed when the processing of S3 to S8 is completed for all the sets. (Step S9), and the loop ends (Step S10).

Assume that when the user 3 is assigned to the first transmission stream, the user 2 is assigned to the second transmission stream, and the user 1 is assigned to the third transmission stream, the evaluation values di and max are the smallest. At this time, the user allocation unit 102 is controlled by the scheduling unit 101, and exchanges the transmission stream s, that is, changes the allocation of each user to each transmission stream. The vector of the transmission stream after the replacement is s ′ = P × s = [s′3, s′2, s′1] ^T in this example.

Next, the feedback processing / modulo operation unit performs processing using the weight stored in step S8. The output of the feedback processing / modulo operation unit (the signal after the modulo operation) v ′ after performing scheduling in this way is expressed as v ′ = B ′ ⁻¹ (by using the weight B ′ and the perturbation vector d ′ after scheduling. s '+ d'). The signal v ′ after this modulo calculation is sent to the beamformer 106 and processed using the weight Q ′ stored in step S8. The output of the beamformer 106 is x ′ = Q′v ′ = B ′ ⁻¹ (s ′ + d ′).

On the other hand, the received signal in the radio receiving apparatus of FIG. 2 can be written as follows by replacing equation (16) with the signal after scheduling.

  At this time, the perturbation vectors d′ 1, d′ 2, and d′ 3, which are signal components added to the reception signal by modulo calculation, are minimized by the processing of FIG. 9, and thus the dynamic range of the reception signal y ′. Is minimized. Therefore, the burden on the analog circuit (for example, the reception RF unit 302) and the digital circuit (ADC 303, FFT unit 304, etc.) of the radio reception apparatus is reduced, and an inexpensive radio reception apparatus can be realized.

(Second Embodiment)
Next, a second embodiment of the present invention will be described. In the present embodiment, a technique for performing transmission with the best characteristics on the premise of using an inexpensive radio reception apparatus having a small dynamic range (referred to as reception dynamic range) of a receivable signal will be described. In the present embodiment, scheduling is performed so that good characteristics can be obtained when the dynamic range of the received signal is limited to a predetermined value.

Now, for example, it is assumed that the absolute value of the reception dynamic range of the wireless reception device of user 3 is V3, max. In this case, s3 + d3 can be obtained only at V3, max among the reception signals of the wireless reception device of user 3 shown in Expression (16). In other words, in general, when there is a restriction that the reception dynamic range of the i-th wireless reception device is Vi, max,

Therefore, the perturbation vector is also limited,

It becomes. In addition,

far. In other words, when the reception dynamic range is limited, the output of the feedback processing / modulo arithmetic unit may not be within ± √M / m.

  Therefore, the number of outputs of the feedback processing / modulo arithmetic unit when the reception dynamic range is limited is calculated.

The maximum value of the signal before adding the perturbation vector is

Therefore, the maximum value of the signal v after adding a limited perturbation vector is

It becomes. However, min (a, b) is defined as a function that outputs the smaller of a and b.

  That is, when the restricted perturbation vector is sufficiently larger than sk, max ′ subjected to feedback processing and modulo arithmetic, vi and max should be suppressed to within ± √M / m, as in the first embodiment. Can do. On the other hand, when the limited perturbation vector is not sufficiently large, the minimum value | sk, max'-di, max | is within the constraints. When vi, max exceeds ± √M / m, the transmission apparatus performs transmission with the power reduced as a whole, and thus the reception apparatus loses reception power. That is, the smaller vi and max are, the better. Therefore, when the reception dynamic range of the receiving device is limited, vi and max are calculated within the limitation, and transmission stream allocation may be performed so that vi and max can be minimized. The above is the principle of the second embodiment.

  The configuration of the wireless transmission device according to the second embodiment is the same as that of the first embodiment shown in FIG. Next, scheduling will be described in detail with reference to FIG. In FIG. 9, steps S1 to S4 are the same as those in the first embodiment, and thus description thereof is omitted.

  In the present embodiment, the evaluation value calculated in step S5 is vi, max shown in Expression (32). For vi and max, only the value of the lowest transmission stream may be used, the sum of vi and max of all transmission streams may be used, or these vi and max may be used by weighted synthesis. good.

  Next, the calculated evaluation value is compared with a threshold value (step S6). As the threshold value, an evaluation value in a combination calculated in the past is used. Since di, max should be smaller, if the evaluation value (threshold value) in the previously calculated combination is better than the evaluation value in the currently calculated combination, the process returns to step S3 and proceeds to the next combination. .

  If the evaluation value in the combination currently calculated is better than the evaluation value (threshold value) in the combination calculated in the past, the current evaluation value is stored (step S7). Since the subsequent processing is the same as that of the first embodiment, description thereof is omitted.

  As described above, in the second embodiment, an allocation method that provides the best characteristics is determined within a predetermined reception dynamic range limit of the reception device. Therefore, even if an inexpensive reception device is used, good communication quality is achieved. It becomes possible to provide.

  In the present embodiment, since the information on the absolute value Vi, max of the reception dynamic range of each user's receiving device is required by the transmitting device, it is necessary to notify the transmitting device of the value of Vi, max using some method. is there.

  First, a method is conceivable in which the Vi and max values are quantized and transmitted as control information such as 4 bits. In this case, the user allocation unit 102 is controlled by the scheduling unit 101, and the signal component added by the modulo arithmetic unit 104 so that the dynamic range of the reception signal of the reception device falls within the reception dynamic range of the reception device according to this control information. Assign transmission streams to limit.

  As another method, the modulo arithmetic unit 104 may transmit control information indicating a limit of a numerical value (number of additions / subtractions) γ indicating how many times a perturbation vector that is a signal component accompanying the modulo arithmetic can be added or subtracted. For example, in the case of QPSK with γ = 3, γ indicates whether the value of N = 2√M / m = 2√2 can be added to or subtracted from the I phase and the Q phase up to three times. In this case, the user allocation unit 102 is controlled by the scheduling unit 101, and the addition / subtraction of the perturbation vector in the modulo arithmetic unit 104 is performed so that the dynamic range of the received signal of the receiving device falls within the reception dynamic range according to the control information indicating the addition / subtraction number γ. Transmission stream allocation is performed so that the limit of the number of times is satisfied.

(Third embodiment)
Next, a third embodiment of the present invention will be described. In the third embodiment, when there are more users (wireless reception devices) than the number of users that can be multiplexed by the wireless transmission device, an optimum user that can be multiplexed is extracted from the users, and the first user is further selected from the users. Alternatively, scheduling is performed based on the second embodiment.

  FIG. 10 shows a wireless transmission device according to the third embodiment. The only difference from the first and second embodiments is the number of users that can be multiplexed in the user allocation unit 102 by the transmission device (here, 3). ) The data signals D1 to D5 to be transmitted to the above users are input.

  The operation of this embodiment follows the flowchart shown in FIG. 9, and the only difference from the first embodiment is the difference in the number of combinations of users and transmission streams. In the present embodiment, the optimum three users are selected from five users for transmission. In this case, the number of combinations is 5 × 4 × 3 = 60. For these 60 combinations, the processing from step S1 to S9 is performed. The data signals for the optimum three users selected based on this processing are allocated to the transmission stream by the user allocation unit 102, and transmission is performed.

  Here, as the evaluation value in the scheduling in the third embodiment, a combination that minimizes the dynamic range of the received signal can be selected as in the first embodiment, or as in the second embodiment. It is also possible to select an evaluation value with the best reception characteristics after limiting the reception dynamic range of the reception apparatus.

(Fourth embodiment)
Next, a fourth embodiment of the present invention will be described. The configuration of the wireless transmission device according to the fourth embodiment is the same as that in FIG. The fourth embodiment is almost the same as the third embodiment, but the process flow in scheduling is different.

  The processing procedure of this embodiment will be described with reference to the flowchart of FIG. In the present embodiment, from five users, first, the scheduling unit 101 selects three users to be accommodated according to the first rule, and then the user assignment unit 102 assigns the three users to the transmission stream according to the second rule. . Here, the norm used in the first embodiment or the second embodiment can be used as the first norm. Various methods can be applied to the first rule, but an example using the Greedy algorithm will be described here. Details of the Greedy algorithm are described in “Multiuser Diversity for a Dirty Paper Approach” by Z. Tu et al.

  First, a loop for selecting three users from five users using the Greedy algorithm is started (step S11). Subsequently, an evaluation value for the dynamic range of the received signal of the receiving device is calculated using the Greedy algorithm (step S12). As the evaluation value, for example, received power when using a feed-forward type weight in which the beam former 106 directs null toward an already selected user is used. When the first user is selected, there is no user already selected, so the weight is the maximum ratio combining weight. Here, it is assumed that the fifth user is selected (step S13).

  Next, it is determined whether the number of accommodated users has been reached (step S14). In this example, since only one user is determined as the wireless transmission device, the process returns to step S12 and loop processing is performed. When accommodating the second user, the evaluation value of the dynamic range of the received signal, that is, the user's Received power is calculated (step S12). Based on this, the user with the largest received power is selected (step S13). In this example, it is assumed that the third user is selected. Similarly, assume that the first user is selected in the next loop.

  When the first, third, and fifth users are selected as described above, the determination to reach the accommodated user is made in step S14, so the loop for user selection is terminated (step S15), and the accommodated user is determined. (Step S16).

  Next, scheduling is performed in consideration of the reception dynamic range of the receiving apparatus for the first, third, and fifth users selected this time. Since the processing from step S17 to S25 is the same as the processing from step S2 to S10 shown in FIG. 9 of the first and second embodiments, description thereof will be omitted.

  As described above, in the present embodiment, in the multi-user THP, the accommodated user is determined using the Greedy algorithm with a relatively small amount of calculation, and thereafter, the transmission stream allocation within the accommodated user is performed in consideration of the reception dynamic range of the receiving apparatus. It can be performed. Therefore, it is possible to provide a simple wireless receiving device while maintaining good communication quality while reducing the amount of calculation of the wireless transmitting device.

  Although the present embodiment has been described assuming a single carrier method, the present invention can also be implemented using a multicarrier method. In this embodiment, the example in which the channel information is obtained by feedback from the receiving device has been described. However, for the channel information, it is also possible to use the channel information estimated when the radio having the transmitting device performs reception. is there.

  Further, in the present embodiment, a dynamic range reduction method mainly using Tomlinson-Harashima Precoding or a method for obtaining the best characteristics among the reception dynamic range restrictions of the reception apparatus has been described. The present invention is also applicable to the case of using the perturbation vector calculation method and beamforming method shown in “A vector-perturbation technique for near-capacity multiantenna multiuser communication \ -Part II: Perturbation” by Peel et al.

  The present invention is not limited to the above-described embodiments as they are, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

1 is a block diagram showing a wireless transmission device according to an embodiment of the present invention. The block diagram which shows the radio | wireless receiver according to the embodiment The block diagram which shows the detail of the channel estimation part in FIG. Diagram explaining the effect of beamforming Diagram explaining the effects of precoding and beamforming The figure which shows the constellation of the transmission signal point on IQ plane for demonstrating the embodiment The figure which shows the constellation of the received signal point on IQ plane for demonstrating the embodiment Diagram showing the relationship between the signal after feedback processing and the perturbation vector Flowchart for explaining transmission processing in the embodiment The block diagram which shows the wireless transmitter according to other embodiment of this invention Flowchart for explaining transmission processing in the embodiment

Explanation of symbols

100: Weight calculation unit 101 ... Scheduling unit (evaluation value calculation unit)
DESCRIPTION OF SYMBOLS 102 ... User allocation part 103 ... Subtraction part 104 ... Modulo calculation part 105 ... Feedback part 106 ... Beam former 107 ... IFFT unit 108 ... Digital-analog converter 109-1 ˜109-n—Transmission RF unit 110-1 to 110-n—Transmission antenna 301—Reception antenna 302—Reception RF unit 303—Analog-digital converter 304—FFT unit 305: Modulo calculation unit 306 ... Channel estimation unit 307 ... Demodulation unit

Claims (12)

In a radio transmission method for transmitting a transmission signal to a plurality of radio receiving apparatuses at the same frequency and at the same time using a plurality of antennas,
Adding a perturbation vector so that the transmitted signal is within a predetermined level to reduce the increase in transmit power;
The signal to be transmitted is at least one transmission so that the dynamic range of the receiver that receives the transmission signal is minimized or the dynamic range of the reception signal is within the reception dynamic range of the radio receiver. Assigning to a stream. A wireless transmission method comprising:   The wireless transmission method according to claim 1, wherein the step of adding the perturbation vector uses Tomlinson-Harashima Precoding. Performing a feedback process of multiplying a signal to be transmitted by a gain, feeding back a signal that causes interference in a wireless reception device, and subtracting the signal that causes interference from the signal to be transmitted;
Performing a calculation for reducing an increase in transmission power due to the feedback processing on the signal after the feedback processing;
Performing beam forming processing for reducing interference remaining in the feedback unit on the signal after the calculation;
Transmitting a transmission signal including the signal after the beamforming process;
Calculating an evaluation value representing a dynamic range of a reception signal in a wireless reception device that receives the transmission signal;
Based on the evaluation value, the signal to be transmitted is at least one transmission stream so that the dynamic range of the reception signal is minimized or the dynamic range of the reception signal is within the reception dynamic range of the radio reception apparatus. Assigning to the wireless transmission method. A feedback processing unit that performs a feedback process of multiplying a signal to be transmitted by a gain, feeding back a signal that causes interference in a wireless reception device, and subtracting the signal that causes interference from the signal to be transmitted;
An arithmetic unit that performs an operation to reduce an increase in transmission power due to the feedback processing on the signal after the feedback processing;
A beam former for performing beam forming processing for reducing interference remaining in the feedback unit with respect to the signal after the calculation;
A transmission unit for transmitting a transmission signal including the signal after the beam forming process;
A calculation unit that calculates an evaluation value representing a dynamic range of a reception signal in a wireless reception device that receives the transmission signal;
Based on the evaluation value, the signal to be transmitted is at least one transmission stream so that the dynamic range of the reception signal is minimized or the dynamic range of the reception signal is within the reception dynamic range of the radio reception apparatus. And a allocating unit for allocating to the wireless transmission device.   The radio transmission apparatus according to claim 4, wherein the transmission unit transmits the transmission signal using a plurality of antennas. The calculation unit calculates the evaluation value for all combinations of a plurality of transmission streams and a plurality of wireless reception devices,
The radio transmission apparatus according to claim 4, wherein the allocation unit determines one combination that minimizes the evaluation value, and performs the allocation based on the determined combination.   The allocation unit performs the allocation to limit a signal component added by the arithmetic unit so that a dynamic range of the reception signal falls within a reception dynamic range of the radio reception apparatus. 5. The wireless transmission device according to 4.   The radio transmission apparatus according to claim 4, wherein the allocating unit receives information indicating the reception dynamic range from the radio reception apparatus.   The said allocation part receives the information which shows the restriction | limiting of the frequency | count of addition / subtraction of the signal component added by the said calculation from the said radio | wireless receiving apparatus, The said allocation is performed so that this restriction | limiting may be satisfy | filled. Wireless transmitter. The calculation unit calculates the evaluation value for all combinations of a plurality of transmission streams and m (plurality) of wireless reception devices,
The allocating unit selects a combination of the plurality of transmission streams and n (n <m) radio receiving apparatuses in ascending order of the evaluation value, and performs the allocation based on the selected combination. The wireless transmission device according to claim 4.   means for selecting n (n <m) radio receivers using a standard from among m (plural) radio receivers, wherein the calculating unit and the assigning unit are configured to receive the selected radio receivers; 5. The wireless transmission device according to claim 4, wherein the evaluation value is calculated and assigned in correspondence with the device.   The radio transmission apparatus according to claim 11, wherein a Greedy algorithm is used as the standard.

Images