JP2008228145A - Receiver, control program, and reception control method in multi-user MIMO system - Google Patents

Abstract

A receiver, a control program, and a reception control method in a multi-user MIMO system capable of reducing the amount of calculation based on a QRM-MLD method for signal separation.
The determination table includes, for each set (MCS) including a modulation scheme and a coding rate, a relative difference between a signal-to-noise ratio of a reference set and a signal-to-noise ratio of the set, and the set for a bit error rate. Are associated with the slope of the signal-to-noise ratio characteristic curve. The receiver uses the decision table to derive a relative signal-to-noise ratio based on the signal-to-noise ratio and the relative difference for each received signal, and the relative signal-to-noise ratio includes a bit error in the reference set. The relative signal-to-noise ratio is derived by multiplying the relative slope ratio of the slope of the bit error rate characteristic curve of the set with respect to the slope of the rate characteristic curve. And the order of the channel matrix H is notified to the QR decomposition | disassembly means in an order from the received signal with a large relative signal-to-noise ratio.
[Selection] Figure 2

Description

  The present invention relates to a receiver, a control program, and a reception control method in a multi-user MIMO (Multi Input Multi Output) system.

  FIG. 1 is a configuration diagram of a MIMO system in the prior art.

Ｎｔ：送信アンテナ数（図１によれば４）
Ｎｒ：受信アンテナ数（図１によれば４）
Ｈ：送受信アンテナ間のチャンネル行列
ｓ（ｔ）：送信シンボル
ｎ（ｔ）：受信機雑音
ｈ_ｉｊ：送信アンテナｊと受信アンテナｉの間のチャネル減衰 According to the MIMO system, different signals are simultaneously transmitted from a plurality of transmission antennas at the same frequency. These transmission signals are received by a plurality of receiving antennas. The received signal is expressed by the following equation.

Nt: Number of transmitting antennas (4 according to FIG. 1)
Nr: Number of receiving antennas (4 according to FIG. 1)
H: Channel matrix between transmission / reception antennas s (t): Transmission symbol n (t): Receiver noise h _ij : Channel attenuation between transmission antenna j and reception antenna i

  The receiver performs processing of separating a plurality of transmission signals from the transmitter into individual signals based on the reception signals received by the respective reception antennas.

  One of the signal separation methods in the receiver is an MLD (Maximum Likelihood Detection) method. This calculates the Euclidean distance or its square for all possible combinations of a plurality of transmission signals and reception signals transmitted from a plurality of transmission antennas. Then, the combination of transmission signals giving the minimum distance is selected as the most probable estimation result. According to this method, it is possible to reliably separate a plurality of transmission signals into individual signals, but due to the calculation of the square Euclidean distance many times, the processing load required for signal separation becomes very large. There is a problem.

  For example, in the case of 16QAM (Quadrature Amplitude Modulation), there are 16 candidates for each transmission signal. In this case, in order to separate the four transmission signals, it is necessary to calculate 16 × 16 × 16 × 16 = 65536 Euclidean distances.

  As a signal separation method improved from the MLD method, there is a QRM-MLD method. This can reduce the number of squared Euclidean distance calculations by improving the MLD method using QR decomposition and the M algorithm. In the QRM-MLD method, not all transmission symbols are estimated at once, but estimation is performed while narrowing down candidates one by one from transmission symbols with a low error probability. This reduces the number of calculations of the Euclidean distance.

  FIG. 1 shows a functional configuration of a receiver based on the QR-MLD method.

  The receiver based on QR-MLD includes a QR decomposition unit 11 that performs QR decomposition on the channel matrix H, a multiplexing unit 12 that multiplies the received signal by Hermitian transposition of the Q matrix and orthogonalizes the received signal, and a branch metric for each stage. And a plurality of surviving symbol replica candidate control units 13 for controlling the number of surviving symbol replica candidates. The surviving symbol replica candidate control unit 13 adaptively selects only symbol replica candidates having a cumulative branch metric smaller than a threshold controlled on the basis of the minimum value of the cumulative branch metric and the noise power as surviving candidates.

<Step 1>
Using the channel matrix H, a received SINR (Signal-to-Interference. And Noise power Ratio) is calculated. Based on the result, the columns of the transmission signal and H are sorted in ascending order.
_{^{SINRp = Σ Nr q = 1 |}} h qp | 2
p: Transmitting antenna [1, Nt]
q: Receive antenna [1, Nq]

ＳＩＮＲ_１＝|ｈ₁₁|^２＋|ｈ₂₁|^２＋|ｈ₃₁|^２＋|ｈ₄₁|^２
ＳＩＮＲ_２＝|ｈ₁₂|^２＋|ｈ₂₂|^２＋|ｈ₃₂|^２＋|ｈ₄₂|^２
ＳＩＮＲ_３＝|ｈ₁₃|^２＋|ｈ₂₃|^２＋|ｈ₃₃|^２＋|ｈ₄₃|^２
ＳＩＮＲ_４＝|ｈ₁₄|^２＋|ｈ₂₄|^２＋|ｈ₃₄|^２＋|ｈ₄₄|^２ For example, in the case of a 4 × 4 system as shown in FIG. 1, it is expressed as follows.

SINR ₁ = | h ₁₁ | ² + | h ₂₁ | ² + | h ₃₁ | ² + | h ₄₁ | ²
SINR ₂ = | h ₁₂ | ² + | h ₂₂ | ² + | h ₃₂ | ² + | h ₄₂ | ²
SINR ₃ = | h ₁₃ | ² + | h ₂₃ | ² + | h ₃₃ | ² + | h ₄₃ | ²
SINR ₄ = | h ₁₄ | ² + | h ₂₄ | ² + | h ₃₄ | ² + | h ₄₄ | ²

Each column of the channel matrix H and each element of the transmission symbol vector s (t) are sorted in ascending order of reception SINR.
H ~: Sorted channel matrix s ~ (t): Sorted transmission symbol

ｓ~(t)＝［ｓ３，ｓ２，ｓ１，ｓ４］^Ｔ For example, when SINR3 <SINR2 <SINR1 <SINR4, it is expressed as follows.

s ~ (t) = [s3, s2, s1, s4] ^T

The received signal z (t) is expressed by the following equation.
z (t) = H ~ .s ~ (t) + n (t) Equation (1)
z (t) and y (t) differ only in order.

  The estimation for a signal transmitted from a transmission antenna having a high SINR has good accuracy. Therefore, candidate signals are first narrowed down by estimating a highly accurate signal. Thereby, the calculation amount can be reduced as a whole.

<Step 2>
Next, QR decomposition is performed on the sorted channel matrix. Then, a unitary matrix Q that satisfies H˜ = Q · R and an upper triangular matrix R are derived. The unitary matrix Q in this case satisfies Q ^H Q = QQ ^H = I. Superscript H represents conjugate transpose, and I represents a unit matrix. The upper triangular matrix R is represented by the following equation.

Equation (1) can be rewritten as follows:
Q ^H y (t) = Q ^H QRs˜ (t) + Q ^H n (t)
z (t) = Q ^H y (t) = Rs˜ (t) + n ′ (t)

Therefore, the above relational expression is expressed as follows.
z ₁ = r ₁₁ y ₁ + r ₁₂ y ₂ + r ₁₃ y ₃ + r ₁₄ y ₄
z ₂ = r ₂₂ y ₂ + r ₂₃ y ₃ + r ₂₄ y ₄
z ₃ = r ₃₃ y ₃ + r ₃₄ y ₄
z ₄ = r ₄₄ y ₄

<Step 3>
Maximum likelihood determination is performed in the order of z ₄ (s ₄ ) → z ₃ (s ₁ ) → z ₂ (s ₂ ) → z ₁ (s ₃ ).
(1) Attention is paid to the equation regarding z ₄ . Matrix element r ₄₄ is known and z ₄ does not interfere with other signals. Then, it can be seen that relies only on a single transmission signal s _4. Therefore, the transmission signal s ₄ has only signal point candidates according to the signal constellation. That is, there are 4 for QPSK and 16 for 16QAM. The candidate set of signal points is represented by C, and the square Euclidean distance is calculated by the following equation.
e ₁ = | z ₄ (t) −r ₄₄ C | ²
Thereby, S1 candidates are selected in ascending order of distance, and they become survival candidates.
S1 ≦ size (C) Number of Euclidean distance calculations: Size (C)

(2) Next, pay attention to the equation regarding z ₃ . The matrix elements r ₃₃ and r ₃₄ are known, x4 has S1 candidates, and x3 has signal point candidates according to the signal constellation. Therefore, there can be combinations of signal points as S1 × Size (C). The square Euclidean distance is calculated S1 × Size (C) times, and the candidates are narrowed down by selecting the combination of S2 (S2 <S1 * Size (C)) in ascending order of the distance.

(3) Next, attention is paid to the equation regarding z2. The matrix elements r22, r23, and r24 are known, and the combinations of the transmission signals x3 and x4 are narrowed down to S2 candidates in the previous stage, and there are Size (C) pieces of x2. Therefore, there can be combinations of signal points as S2 × Size (C). The square Euclidean distance is calculated S2 * Size (C) times, and candidates are narrowed down by selecting a combination of S3 (S3 <S2 * Size (C)) in ascending order of the distance.

(4) Finally, attention is paid to the expression relating to z1. The matrix elements r11, r12, r13, and r14 are known, and the combinations of the transmission signals x2, x3, and x4 are narrowed down to S3 candidates in the previous stage, and size (C) signal point candidates for x1 Exists. Therefore, there can be combinations of signal points as S3 × Size (C). The candidates are narrowed down by calculating the square Euclidean distance in S3 * Size (C) times and selecting the combination of S4 in ascending order of the distance.

JP 2006-287613 A

  The MIMO system can support multiple users. That is, a signal is simultaneously transmitted from a plurality of transmitters having one transmission antenna to a receiver having a plurality of reception antennas using the same frequency resource.

  According to the multiuser MIMO system, different signals are transmitted from a plurality of transmission antennas to a receiver using the same frequency resource at the same time. The receiver needs to separate a plurality of transmission signals from each transmission antenna into individual signals based on the reception signal received by each reception antenna.

  When QRM-MLD is applied to a multi-user MIMO system, first, an error rate of transmission data from each transmission antenna is estimated, and sorting is performed so as to estimate sequentially from transmission signals of transmission antennas having a low error rate. There is a need to.

  According to the prior art, received SINR is used as an index for sorting. In a single user MIMO system, the transmission data modulation scheme of all transmission antennas and the coding rate of the error correction code in the physical layer are the same, so the higher the received SINR, the lower the error rate. Can do.

  However, in a multiuser MIMO system, the modulation scheme of the physical layer and the coding rate of the error correction code are different for each transmission antenna, and therefore the error rate of the transmission antenna cannot be estimated only from the reception SINR.

  Assuming that the same coding rate is used, even if the received SINR is low, the error rate of the low-speed modulation scheme may be lower than the error rate of the high-speed modulation scheme. That is, when determining whether the error rate is high or low, it is necessary to consider not only the received SINR but also the modulation scheme and the coding rate of the error correction code. Even if the error rate is the same, it is preferable to estimate the transmission signal in order from the low-speed modulation scheme to the high-speed modulation scheme in order to reduce the number of calculations of the Euclidean distance.

  Therefore, an object of the present invention is to provide a receiver, a control program, and a reception control method in a multi-user MIMO system that can reduce the amount of calculation based on the QRM-MLD method for signal separation.

According to the present invention, a receiver in a multi-user MIMO system, QR decomposition means for QR-decomposing a channel matrix H, multiplexing means for multiplying a received signal by Hermitian transpose of the Q matrix and orthogonalizing, a stage In each receiver having a plurality of surviving symbol replica candidate control means for controlling the number of surviving symbol replica candidates,
For each set of modulation scheme and coding rate, the relative difference between the signal-to-noise ratio of one reference set and the signal-to-noise ratio of the set at the target error rate, and the signal-to-noise ratio of the set with respect to the bit error rate A determination table that correlates the slope of the characteristic curve;
Using a determination table, for each received signal, a relative signal-to-noise ratio based on the signal-to-noise ratio and the relative difference of the received signal is derived, and the bit error rate of the reference set is calculated as the relative signal-to-noise ratio. A relative signal-to-noise ratio calculating means for deriving a relative signal-to-noise ratio obtained by multiplying a relative slope ratio of the slope of the bit error rate characteristic curve of the set with respect to the slope of the characteristic curve;
Estimation order determining means for notifying the QR decomposition means of the order of the channel matrix H so as to be estimated by the multiplexing means and the symbol replica candidate control means in order from the received signal having the larger relative signal-to-noise ratio. Features.

  According to another embodiment of the receiver of the present invention, it is also preferable that the estimation order determining means rearranges the channel matrix H in the order of the low-speed modulation scheme to the high-speed modulation scheme when the relative signal-to-noise ratio is the same.

  According to another embodiment of the receiver of the present invention, the signal to noise ratio is also preferably SINR.

According to another embodiment of the receiver of the present invention,
The signal to noise ratio is Eb / No (power density per bit to noise power density ratio),
A power density ratio calculating means for calculating Eb / No from the SINR of the received signal;
The determination table associates, for each set, the relative difference between Eb / No of one reference set and Eb / No of the set, and the slope of the Eb / No characteristic curve of the set with respect to the bit error rate,
The relative signal-to-noise ratio calculating means derives the relative Eb / No for each received signal based on the Eb / No and the relative difference of the received signal using the determination table, and the relative Eb / No And a relative signal-to-noise ratio obtained by multiplying the slope of the bit error rate characteristic curve of the reference set with the slope of the bit error rate characteristic curve of the set.
It is also preferable that the estimation order determining means determines the order of the channel matrix H in order from the received signal having a larger relative Eb / No.

According to another embodiment of the receiver of the present invention,
A plurality of judgment tables are provided based on the fading speed and / or delay dispersion value,
Radio wave environment estimating means for estimating a fading speed and / or delay dispersion value from a received signal;
It is also preferable to have determination table selection means for selecting a determination table based on the fading speed and / or delay dispersion value.

According to the present invention, a receiver in a multi-user MIMO system, QR decomposition means for QR-decomposing a channel matrix H, multiplexing means for multiplying a received signal by Hermitian transpose of the Q matrix and orthogonalizing, a stage In each control program for functioning a computer mounted on a receiver having a plurality of surviving symbol replica candidate control means for controlling the number of surviving symbol replica candidates,
For each set of modulation scheme and coding rate, the relative difference between the signal-to-noise ratio of one reference set and the signal-to-noise ratio of the set at the target error rate, and the signal-to-noise ratio of the set with respect to the bit error rate A determination table that correlates the slope of the characteristic curve;
Using a determination table, for each received signal, a relative signal-to-noise ratio based on the signal-to-noise ratio and the relative difference of the received signal is derived, and the bit error rate of the reference set is calculated as the relative signal-to-noise ratio. A relative signal-to-noise ratio calculating means for deriving a relative signal-to-noise ratio obtained by multiplying a relative slope ratio of the slope of the bit error rate characteristic curve of the set with respect to the slope of the characteristic curve;
The computer functions as an estimation order determining unit that notifies the QR decomposition unit of the order of the channel matrix H so as to be estimated by the multiplexing unit and the symbol replica candidate control unit in order from the received signal having the larger relative signal-to-noise ratio. It is characterized by that.

According to another embodiment of the control program for a receiver of the present invention,
The estimation order determination means preferably causes the computer to function so as to rearrange the channel matrix H in the order from the low-speed modulation method to the high-speed modulation method when the relative signal-to-noise ratio is the same.

  According to another embodiment of the control program for a receiver of the present invention, it is also preferable to make the computer function so that the signal-to-noise ratio is SINR.

According to another embodiment of the control program for a receiver of the present invention,
The signal to noise ratio is Eb / No (power density per bit to noise power density ratio),
A power density ratio calculating means for calculating Eb / No from the SINR of the received signal;
The determination table associates, for each set, the relative difference between Eb / No of one reference set and Eb / No of the set, and the slope of the Eb / No characteristic curve of the set with respect to the bit error rate,
The relative signal-to-noise ratio calculating means derives the relative Eb / No for each received signal based on the Eb / No and the relative difference of the received signal using the determination table, and the relative Eb / No And a relative signal-to-noise ratio obtained by multiplying the slope of the bit error rate characteristic curve of the reference set with the slope of the bit error rate characteristic curve of the set.
The estimation order determination means preferably causes the computer to function so as to determine the order of the channel matrix H in order from a received signal having a larger relative Eb / No.

According to another embodiment of the control program for a receiver of the present invention,
A plurality of judgment tables are provided based on the fading speed and / or delay dispersion value,
Radio wave environment estimating means for estimating a fading speed and / or delay dispersion value from a received signal;
It is also preferable to cause the computer to function as a determination table selection unit that selects a determination table based on the fading speed and / or the delay dispersion value.

According to the present invention, there is provided a reception control method for a receiver in a multi-user MIMO system, comprising:
For each set of modulation scheme and coding rate, the relative difference between the signal-to-noise ratio of one reference set and the signal-to-noise ratio of the set at the target error rate, and the signal-to-noise ratio of the set with respect to the bit error rate It has a judgment table that correlates the slope of the characteristic curve,
Using a determination table, for each received signal, a relative signal-to-noise ratio based on the signal-to-noise ratio and the relative difference of the received signal is derived, and the bit error rate of the reference set is calculated as the relative signal-to-noise ratio. A first step of deriving a relative signal to noise ratio multiplied by a relative slope ratio of the slope of the bit error rate characteristic curve of the set to the slope of the characteristic curve;
A second step of notifying the QR decomposition unit of the order of the channel matrix H so as to be estimated by the multiplexing unit and the symbol replica candidate control unit in order from the received signal having a large relative signal-to-noise ratio;
A third step of QR decomposition based on the channel matrix H;
And a fourth step of controlling the number of surviving symbol replica candidates for each stage.

  According to the receiver, control program, and reception control method of the multiuser MIMO system of the present invention, the relative signal-to-noise ratio is calculated for each set of modulation scheme and coding rate, and the channel matrix is calculated in descending order of the ratio. By rearranging H, the amount of calculation based on the QRM-MLD method for signal separation can be reduced.

  Hereinafter, the best mode for carrying out the present invention will be described in detail with reference to the drawings.

  FIG. 2 is a functional configuration diagram of the multi-user MIMO receiver according to the present invention.

  According to FIG. 2, a receiver 1 for multi-user MIMO is shown, using the same frequency resource from a plurality of transmitters having one transmission antenna to a receiver having a plurality of reception antennas. Send a signal at the same time.

  Compared to FIG. 1, the receiver 1 of FIG. 2 includes a channel estimation unit 14, a reception SINR estimation unit 15, a power density ratio calculation unit 16, a determination table unit 17, and a relative signal-to-noise ratio calculation unit. 18, an estimation order determination unit 19, a radio wave environment estimation unit 1 </ b> A, and a determination table selection unit 1 </ b> B. These functional configurations can also be realized by executing a control program that causes a computer mounted on the receiver to function. For example, it can be realized by an FPGA (Field Programmable Gate Array) or DSP (Digital Signal Processor) which is a programmable processor.

  The reception SINR estimation unit 15 estimates the reception SINR for each transmission signal. The estimated received SINR is output to the relative signal-to-noise ratio calculation unit 18 or the power density ratio calculation unit 16.

  The determination table unit 17 calculates the signal-to-noise ratio of 11 reference sets at the target error rate for each set (MCS: Modulation and Coding rate Set) consisting of a modulation scheme and a coding rate. A determination table in which a relative difference with the signal-to-noise ratio of the set is associated is stored. Further, the determination table also associates the slope of the signal-to-noise ratio characteristic curve with respect to the bit error rate for each MCS.

  The signal-to-noise ratio may be received SINR or Eb / No (power density per bit to noise power density ratio).

Table 1 is a determination table when the signal-to-noise ratio is the received SINR.

  FIG. 3 is a graph of the bit error rate with respect to the received SINR when the signal-to-noise ratio is the received SINR.

  According to FIG. 3, a target error rate is set, and a reception SINR that achieves the target error rate is shown for each MCS. Then, with reference to the set “BPSK, 1/2” having the lowest received SINR, the distances to the received SINRs of the respective sets that achieve the target error rate are represented by c1 to c6.

  FIG. 4 is a graph of the bit error rate with respect to Eb / No when the signal-to-noise ratio is Eb / No.

  According to FIG. 4, a target error rate is set, and Eb / No that achieves the target error rate is represented for each MCS. Then, with reference to the set “BPSK, 1/2” having the lowest Eb / No, the distances to Eb / No of each set that achieve the target error rate are represented by a1 to a6.

Table 2 is a determination table when the signal-to-noise ratio is Eb / No.

  3 and 4 are derived in an AWGN (additive white Gaussian noise) environment in a system in which convolutional coding is applied to an error correction code.

The power density ratio calculation unit 16 is provided only when the signal-to-noise ratio is Eb / No, and converts Eb / No in the digital modulation signal from the received SINR.
Received SINR (dB) = Eb / No (dB) + 10 * log10 (M * Coding Rate)
M: Modulation system dependent coefficient
When modulation method = BPSK, M = 2
When modulation method = 16QAM, M = 4
When modulation method = QPSK, M = 2
The calculated Eb / No is notified to the relative signal-to-noise ratio calculation unit 18.

  The relative signal-to-noise ratio calculation unit 18 uses the determination table 17 to derive, for each received signal, a signal-to-noise ratio of the received signal and a relative signal-to-noise ratio based on the relative difference. Then, the relative signal-to-noise ratio is derived by multiplying the relative signal-to-noise ratio by the relative slope ratio of the slope of the bit error rate characteristic curve of the set with respect to the slope of the bit error rate characteristic curve of the reference set.

[When the signal-to-noise ratio is the received SINR]
Relative SINR (dB)
= (Relative signal-to-noise ratio) x (Relative slope ratio)
= {Received SINR-relative difference}
× {Inclination of the BER characteristic curve of the set / Inclination k of the BER characteristic curve of the reference set}

[When the signal-to-noise ratio is Eb / No]
Relative Eb / No (dB)
= (Relative signal-to-noise ratio) x (Relative slope ratio)
= {Eb / No-relative difference}
× {Inclination of the BER characteristic curve of the set / Inclination k of the BER characteristic curve of the reference set}

  Regarding FIG. 3, focus on the first set “QPSK, 2/3” and the second set “QPSK, 3/4” versus the reference set “BPSK, 1/2”.

The received SINR at the target error rate of the first set has a distance (relative difference) of c2 [dB] with respect to the received SINR at the target error rate of the reference set. Here, the following equation is calculated for the actual received SINR.
Received SINR−relative difference c2 = α

The received SINR at the second set target error rate has a distance (relative difference) of c5 [dB] with respect to the received SINR at the reference set target error rate. Here, the following equation is calculated for the actual received SINR.
Received SINR-relative difference c5 = α

  Here, for the first set and the second set, when α (relative signal to noise ratio) matches the target error rate, it is estimated that the error rates of both received signals are equal. However, the set with the lower error rate for the signal-to-noise ratio of the received signal should be selected first.

  According to FIG. 3, for the first set, the error rate in the actual received SINR (SINR + c2 + α of the target error rate of the reference set) is represented as “error rate 1”. For the second set, the error rate at the actual reception SINR (the SINR of the target error rate of the reference set + c5 + α) is expressed as “error rate 2”.

  Here, it is clear that “error rate 2” is lower than “error rate 1”. This state depends on the slope of the curve drawn in the graph of the bit error rate with respect to the received SINR calculated for each set. That is, it is necessary to include not only the relative difference of the received SINR of each set with respect to the reference set but also the slope of the received SINR vs. bit error rate curve of each set with respect to the reference set.

  As another embodiment of the present invention, it is also preferable to provide a plurality of determination tables according to the radio wave environment. The radio wave environment is, for example, fading speed d and / or delay dispersion τ. In a multi-user environment, since fading speed and delay dispersion differ for each user, SINR that can satisfy the target error rate for each MCS and one reference for each fading speed and delay dispersion in order to accurately determine the estimation order. A table of relative differences of MCS sets (for example, BPSK, Coding Rate = 1/2) is prepared.

Table 3 is a plurality of determination tables set when the signal-to-noise ratio is the received SINR.

Table 4 is a plurality of determination tables set when the signal-to-noise ratio is Eb / No.

  The receiver estimates the radio wave environment of each terminal and selects a determination table close to the radio wave environment. Specifically, there are roughly two methods. In the first method, the receiver estimates the radio wave environment of each user using a pre-amp of the received signal. In the second method, each terminal estimates the radio wave environment based on the reception quality of the signal received from the other party (base station), and each terminal transmits the radio wave environment information to the receiver.

As shown in FIG. 2, for example, when there are four transmission antennas (when there are four transmitters), the MCS IDs are “2”, “6”, “3”, and “4”, respectively. Further, it is assumed that antennas 1 and 2 have selected the upper table of Table 3, and antennas 3 and 4 have selected the lower table of Table 4. Further, it is assumed that the relationship of received SINR is SINR3 <SINR1 <SINR2 <SINR4. At this time, the relative SINR is calculated as follows.
Relative SINR1 = (SINR1-c2) × (g2 / k)
Relative SINR2 = (SINR2-c6) × (g6 / k)
Relative SINR3 = (SINR3-d3) × (h3 / k)
Relative SINR4 = (SINR4-d4) × (h4 / k)

Further, it is assumed that the relationship of Eb / No is Eb / No3 <Eb / No1 <Eb / No2 <Eb / No4. At this time, the relative Eb / No is calculated as follows.
Relative Eb / No1 = (SINR1-10 * log10 (2 * 1/2))-a2) * (k2 / k)
Relative Eb / No2 = (SINR2-10 * log10 (4 * 2/3)) − a4) × (k4 / k)
Relative Eb / No3 = (SINR3-10 * log10 (2 * 2/3)) − b3) × (j3 / k)
Relative Eb / No4 = (SINR4-10 * log10 (4 * 1/2)) − a4) × (j4 / k)

  The estimation order determining unit 19 sequentially determines the QR decomposition unit 11 so as to be estimated by the multiplexing unit 12 and the symbol replica candidate control unit 13 in order from a received signal having a larger relative signal-to-noise ratio (received SINR or Eb / No). Is notified of the order of the channel matrix H. Specifically, each column of the channel matrix H and each element of the transmission symbol vector s (t) are sorted in ascending order of relative signal-to-noise ratio. When the relative signal-to-noise ratio is equal, sorting is performed in the order of low-speed modulation method → high-speed modulation method.

  The radio wave environment estimation unit 1A estimates the fading speed and / or delay dispersion value from the received signal. The estimated fading speed and / or delay dispersion value is notified to the determination table selection unit 1B.

  The determination table selection unit 1B selects one determination table from a plurality of determination tables based on the fading speed and / or the delay dispersion value.

  FIG. 5 is a flowchart in the present invention.

(S501) The reception signal output from the channel estimation unit is input.
(S502) One determination table is selected from a plurality of determination tables based on the fading speed and / or the delay dispersion value.
(S503) The received SINR is estimated for each transmission signal.
(S504) If the relative signal-to-noise ratio is Eb / No, the received SINR is converted to Eb / No.
(S505) Using the determination table, for each received signal, a signal-to-noise ratio of the received signal and a relative signal-to-noise ratio based on the relative difference are derived. Then, the relative signal-to-noise ratio is derived by multiplying the relative signal-to-noise ratio by the relative slope ratio of the slope of the bit error rate characteristic curve of the set with respect to the slope of the bit error rate characteristic curve of the reference set.
(S506) For each MCS, the columns of the channel matrix H and the elements of the transmission symbol vector s (t) are sorted in order from the received signal having a smaller relative signal-to-noise ratio. When the relative signal-to-noise ratio is the same, sorting is performed in the order of low-speed modulation method → high-speed modulation method (that is, BPSK → QPSK → 16QAM).
(S507) Each column of the rearranged channel matrix H and each element of the transmission symbol vector s (t) are output to the QR decomposition unit.

  As described above in detail, according to the receiver, control program, and reception control method of the multiuser MIMO system of the present invention, the relative signal-to-noise ratio is calculated for each set of modulation scheme and coding rate. By rearranging the channel matrix H in ascending order of the ratio, the amount of calculation based on the QRM-MLD method for signal separation can be reduced.

  According to the various embodiments of the present invention described above, those skilled in the art can easily make various changes, modifications and omissions within the scope of the technical idea and the viewpoint of the present invention. The above description is merely an example, and is not intended to be restrictive. The invention is limited only as defined in the following claims and the equivalents thereto.

It is a block diagram of the MIMO system in a prior art. It is a functional block diagram in the receiver for multiuser MIMO in this invention. It is a graph of the bit error rate with respect to reception SINR, when making a signal-to-noise ratio into reception SINR. It is a graph of the bit error rate with respect to Eb / No when a signal-to-noise ratio is set to Eb / No. It is a flowchart in this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Receiver 11 QR decomposition part 12 Multiplexing part 13 Surviving symbol replica candidate control part 14 Channel estimation part 15 Reception SINR estimation part 16 Power density ratio calculation part 17 Judgment table part 18 Relative signal-to-noise ratio calculation part 19 Estimation order determination Part 1A Radio wave environment estimation part 1B Judgment table selection part

Claims (11)

A receiver in a multi-user MIMO (Multi Input Multi Output) system, QR decomposition means for QR-decomposing a channel matrix H, multiplexing means for multiplying a received signal by Hermitian transpose of the Q matrix, and a stage, In each receiver having a plurality of surviving symbol replica candidate control means for controlling the number of surviving symbol replica candidates,
For each set of modulation scheme and coding rate, the relative difference between the signal-to-noise ratio of one reference set and the signal-to-noise ratio of the set at the target error rate, and the signal-to-noise ratio of the set with respect to the bit error rate A determination table that correlates the slope of the characteristic curve;
Using the determination table, for each received signal, a relative signal-to-noise ratio based on the signal-to-noise ratio of the received signal and the relative difference is derived, and the relative signal-to-noise ratio is calculated based on the reference set. A relative signal-to-noise ratio calculating means for deriving a relative signal-to-noise ratio multiplied by a relative slope ratio of the slope of the bit error rate characteristic curve of the set with respect to the slope of the bit error rate characteristic curve;
Estimating order determining means for notifying the QR decomposition means of the order of the channel matrix H so as to be estimated by the multiplexing means and the symbol replica candidate control means in order from the received signal having the larger relative signal-to-noise ratio. And a receiver.   2. The receiver according to claim 1, wherein the estimation order determination unit rearranges the channel matrix H in order from a low-speed modulation scheme to a high-speed modulation scheme when the relative signal-to-noise ratio is the same.   3. The receiver according to claim 1, wherein the signal-to-noise ratio is SINR (Signal-to-Interference. And Noise power Ratio). The signal to noise ratio is Eb / No (power density per bit to noise power density ratio);
A power density ratio calculating means for calculating Eb / No from the SINR of the received signal;
The determination table associates, for each set, the relative difference between Eb / No of one reference set and Eb / No of the set, and the slope of the Eb / No characteristic curve of the set with respect to the bit error rate. And
The relative signal-to-noise ratio calculating means derives a relative Eb / No for each received signal based on the Eb / No of the received signal and the relative difference for each received signal using the determination table. Deriving a relative signal-to-noise ratio by multiplying Eb / No by the relative slope ratio of the slope of the bit error rate characteristic curve of the set to the slope of the bit error rate characteristic curve of the reference set;
3. The receiver according to claim 1, wherein the estimation order determining unit determines the order of the channel matrix H in order from the received signal having the larger relative Eb / No. A plurality of the determination tables are provided based on fading speed and / or delay dispersion value,
Radio wave environment estimating means for estimating the fading speed and / or delay dispersion value from the received signal;
6. The receiver according to claim 1, further comprising a determination table selecting unit that selects the determination table based on the fading speed and / or delay dispersion value. A receiver in a multi-user MIMO system, comprising QR decomposition means for QR-decomposing a channel matrix H, multiplexing means for multiplying a received signal by Hermitian transpose of the Q matrix and orthogonalizing, and a surviving symbol replica for each stage In a control program for functioning a computer mounted on a receiver having a plurality of surviving symbol replica candidate control means for controlling the number of candidates,
For each set of modulation scheme and coding rate, the relative difference between the signal-to-noise ratio of one reference set and the signal-to-noise ratio of the set at the target error rate, and the signal-to-noise ratio of the set with respect to the bit error rate A determination table that correlates the slope of the characteristic curve;
Using the determination table, for each received signal, a relative signal-to-noise ratio based on the signal-to-noise ratio of the received signal and the relative difference is derived, and the relative signal-to-noise ratio is calculated based on the reference set. A relative signal-to-noise ratio calculating means for deriving a relative signal-to-noise ratio multiplied by a relative slope ratio of the slope of the bit error rate characteristic curve of the set with respect to the slope of the bit error rate characteristic curve;
Estimating order determining means for notifying the QR decomposition means of the order of the channel matrix H so as to be estimated by the multiplexing means and the symbol replica candidate control means in order from the received signal having the larger relative signal-to-noise ratio. A control program for a receiver, characterized by causing a computer to function as:   The said estimation order determination means makes a computer function so that the said channel matrix H may be rearranged in order from a low-speed modulation system to a high-speed modulation system, when the said relative signal-to-noise ratio is the same. Program for all receivers.   The program for a receiver according to claim 6 or 7, wherein the computer functions so that the signal-to-noise ratio is SINR. The signal to noise ratio is Eb / No (power density per bit to noise power density ratio);
A power density ratio calculating means for calculating Eb / No from the SINR of the received signal;
The determination table associates, for each set, the relative difference between the Eb / No of one reference set and the Eb / No of the set, and the slope of the Eb / No of the set with respect to the bit error rate,
The relative signal-to-noise ratio calculating means derives a relative Eb / No for each received signal based on the Eb / No of the received signal and the relative difference for each received signal using the determination table. Deriving a relative signal-to-noise ratio by multiplying Eb / No by the relative slope ratio of the slope of the bit error rate characteristic curve of the set to the slope of the bit error rate characteristic curve of the reference set;
8. The receiver according to claim 6, wherein the estimation order determination unit causes a computer to function to determine the order of the channel matrix H in order from the received signal having the larger relative Eb / No. Program. A plurality of the determination tables are provided based on fading speed and / or delay dispersion value,
Radio wave environment estimating means for estimating the fading speed and / or delay dispersion value from the received signal;
10. The receiver control according to claim 6, wherein a computer is caused to function as a determination table selection unit that selects the determination table based on the fading speed and / or delay dispersion value. 11. program. A reception control method for a receiver in a multi-user MIMO system, comprising:
For each set of modulation scheme and coding rate, the relative difference between the signal-to-noise ratio of one reference set and the signal-to-noise ratio of the set at the target error rate, and the signal-to-noise ratio of the set with respect to the bit error rate It has a judgment table that correlates the slope of the characteristic curve,
Using the determination table, for each received signal, a relative signal-to-noise ratio based on the signal-to-noise ratio of the received signal and the relative difference is derived, and the relative signal-to-noise ratio is calculated based on the reference set. Deriving a relative signal-to-noise ratio multiplied by the relative slope ratio of the slope of the bit error rate characteristic curve of the set to the slope of the bit error rate characteristic curve;
A second step of notifying the QR decomposition means of the order of the channel matrix H so as to be estimated by the multiplexing means and the symbol replica candidate control means in order from the received signal having a large relative signal-to-noise ratio;
A third step of QR decomposition based on the channel matrix H;
And a fourth step of controlling the number of surviving symbol replica candidates for each stage by multiplying the received signal by Hermitian transpose of the Q matrix to make it orthogonal.

Images