CN102594524B - Orthogonal space-time block code transmission method based on an optimal relay linear weighting matrix - Google Patents

Abstract

The invention discloses an orthogonal space-time block code transmission method based on an optimal relay linear weighting matrix, which solves the problems of high error rate and poor transmission performance in the traditional method. The orthogonal space-time block code transmission method is realized by the following steps: (1) a communication system formed by a source node S, a relay node R and a destination node D is set; (2) at a first time slot, the source node S sends signal vectors x1 and x2, and the relay node R receives data yr1 and yr2; (3) the relay node R carries out singular value decomposition on a channel matrix H1 from the source node S to the relay node R and a channel matrix h2 from the relay node R to the destination node D; (4) the relay node R constructs an optimal linear weighting matrix W; (5) at a second time slot, the relay node R constructs and sends signal vectors xr1 and xr2 to the destination node D; and (6) the destination node D receives data yd1 and yd2 and constructs a judgment vector to decode. The orthogonal space-time block code transmission method disclosed by the invention has the advantages that: since the signal-to-noise ratio of received signals is increased, the system error rate is reduced, and the system transmission performance is improved.

Description

Based on the orthogonal space time block coding transmission method of the linear weighting matrix of optimum relaying

Technical field

The invention belongs to communication technical field, relate to signal transmission in relay cooperative communication system, is the orthogonal space time block coding transmission method based on the linear weighting matrix of optimum relaying specifically, can be used for many antenna relays system.

Background technology

In wireless communication system, multiple-input and multiple-output MIMO antenna configuration can greatly improve reached at the capacity of traditional single-input single-output SISO system, and diversity can be provided, multiplexing waits gain.For example, in mimo system, by adopting vertical demixing time space transmission method, can obtain spatial reuse gain, and while adopting orthogonal space time block coding transmission method, can obtain diversity gain.

But in actual applications, on the one hand, due to volume and the power consumption constraints of communication terminal, mimo system cannot be realized.On the other hand, because cooperation technology can obtain space diversity gain, and can expand the coverage area, therefore the auxiliary cooperation communication system of relaying receives much concern.Relaying subplan can be divided into two large classes: regeneration scheme and non-renewable scheme.In regeneration scheme, the via node signal of receiving of first decoding, then encode, and then send to destination node.And for non-renewable scheme, its complexity is low, the signal that via node only needs linear processing to receive, then be transmitted to destination node.

For the auxiliary vertical demixing time space transmission method of non-regenerative relaying, if via node is only known the channel condition information CSI between source node and it self, just can adopt a kind of simple amplification coefficient to carry out relaying.But, if the CSI information between source node and via node and between via node and destination node is all known, can, by reasonably designing the linear weighted function matrix at via node place, while making systematic function ratio only adopt simple relaying amplification coefficient, there is great raising.

On the other hand, also obtained considerable research for the auxiliary orthogonal space time block coding transmission system of the non-regenerative relaying shown in Fig. 1.In the system shown in Fig. 1, comprise a source node S, a via node R and a destination node D.Wherein all nodes are all furnished with many antennas.Source node S sends space-time block coding data to via node R, and via node R receives the data from source node S, and amplifies processing, is then transmitted to destination node D.Wherein, via node R amplifies when deal with data, if via node R without any CSI information, what it adopted is fixed gain coefficient.But in the time that relaying node R can obtain the CSI information between source node and relaying, it can adopt the auxiliary simple weighted matrix based on gain coefficient of CSI to transmit.Research shows, by carrying out rational power control, adopts the auxiliary orthogonal space time block coding transmission method based on gain coefficient weighting matrix of CSI can obtain the identical performance of same fixed relay gain method.But above these two kinds of methods all do not make full use of linear process ability and the multi-antenna diversity gain of relaying, thus the reduction of system received signal to noise ratio and the decline of bit error rate performance can be caused, thus reduce the transmission performance of system.And, in the time that the CSI information between source node and relaying and relaying and destination node is all known, the optimum linearity weighting matrix at relaying place is unknown, the information at the processing relaying place that therefore the auxiliary orthogonal space time block coding transmission method of current non-regenerative relaying can not be optimum, make error rate of system performance very low, system transmission performance is limited.

Summary of the invention

The object of the invention is to the deficiency for above-mentioned prior art, propose a kind of orthogonal space time block coding transmission method based on the linear weighting matrix of optimum relaying, to improve the signal to noise ratio that receives signal, thereby improve the bit error rate performance of system.

The technical thought that realizes the object of the invention is, by utilizing the CSI information between source node and relaying and relaying and destination node, the optimum linearity weighting matrix of design retransmit, and improve existing orthogonal space time block coding transmission method, thereby the diversity gain that effectively utilizes many antennas of relaying to provide, make destination node can obtain maximum received signal to noise ratio, improve error rate of system performance.Its specific implementation process is as follows:

(1) set a source node S, a via node R and a destination node D, to build the auxiliary orthogonal space time block coding transmission system of non-regenerative relaying, and a transmission cycle of this system is divided into two time slots, each time slot comprises two moment;

(2) source node S transmitted signal in the first time slot, in the first moment, source node S sends primary signal vector x ₁give via node R, in the second moment, source node S sends the signal vector x after space-time block coding ₂give via node R, and x ₁and x ₂meet following form:

$x_1=\left[\begin{array}{c}{m}_1\\ {\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">m</figure-callout>}}_2\end{array}\right]$

$x_2=\left[\begin{array}{c}{-m}_2^{*}\\ {\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">m</figure-callout>}}_1^{*}\end{array}\right]$

Wherein, m ₁for first modulation symbol at source node S place, m ₂for second modulation symbol at source node S place, () ^*represent conjugate operation;

(3) via node R receives data in the first time slot, and in the first moment, first data that via node R receives are y _r1, in the second moment, second data that via node R receives are y _r2;

(4) via node R uses the training sequence estimation technique to obtain the channel matrix H of source node S to via node R ₁, and it is carried out to singular value decomposition:

$H_1={\mathrm{<figure-callout\; id="1"\; label="U"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">U</figure-callout>}}_1\left[\begin{array}{c}{\widehat{\mathrm{\&Lambda;}}}_1\\ 0\end{array}\right]V_1^H={\mathrm{<figure-callout\; id="1"\; label="U"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">U</figure-callout>}}_1{\mathrm{\&Lambda;}}_1{V}_1^H$

Wherein, U ₁it is the channel matrix H that source node S arrives via node R ₁left singular vector,

for the diagonal matrix of (2 × 2) dimension, V ₁it is the channel matrix H that source node S arrives via node R ₁right singular vector, ${\mathrm{\&Lambda;}}_1=\left[\begin{array}{c}{\widehat{\mathrm{\&Lambda;}}}_1\\ 0\end{array}\right]$ For (N _r× 2) matrix of dimension, N _rfor the antenna number at via node R place, () ^hoperate for conjugate transpose;

(5) via node R obtains the channel matrix h of via node R to destination node D by feedback ₂, and it is carried out to singular value decomposition:

${\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2=\left[\begin{array}{cc}{\mathrm{\&lambda;}}_{21}& 0\end{array}\right]V_2^H={\mathrm{\&Lambda;}}_2{V}_2^H$

Wherein, λ ₂₁the channel matrix h of via node R to destination node D ₂non-zero singular value, V ₂the channel matrix h of via node R to destination node D ₂right singular vector, Λ ₂=[λ ₂₁0] be N _rthe row vector of individual element;

(6) via node R arrives the channel matrix H of via node R according to source node S ₁left singular vector U ₁with the channel matrix h of via node R to destination node D ₂right singular vector V ₂, build and make the optimum linearity weighting matrix W of received signal to noise ratio maximum be:

$W=\sqrt{\frac{P_r}{E_S{\mathrm{\&lambda;}}_{11}^2+{\mathrm{\&sigma;}}_r^2}}{\left(V_2\right)}_1{\left(U_1\right)}_1^H$

Wherein, P _rfor the transmitted power of via node R, E _sfor the energy of source node S transmission symbol, λ ₁₁for source node S is to the channel matrix H of via node R ₁singular value,

for noise power, (V ₂) ₁for via node R is to the channel matrix h of destination node D ₂right singular vector V ₂first row element, (U ₁) ₁for source node S is to the channel matrix H of via node R ₁left singular vector U ₁first row element;

(7) in the second time slot, via node R sends data, and in the first moment, first in the optimum linearity weighting matrix W that via node R constructs step (6) and step (3) receives data y _r1multiply each other, obtain first transmission signal vector of relaying x _r1, and send to destination node D; In the second moment, second in the optimum linearity weighting matrix W that via node R constructs step (6) and step (3) receives data y _r2multiply each other, obtain second transmission signal vector x of relaying _r2, and send to destination node D;

(8) destination node D receives data in the second time slot, and in the first moment, it is y that destination node D receives first data _d1, in the second moment, it is y that destination node D receives second data _d2, and according to receive two data y _d1and y _d2structure judgement vector, to carry out decoding.

Tool of the present invention has the following advantages:

1) the present invention is by utilizing the channel condition information CSI between source node and via node and between via node and destination node, construct the optimum linearity weighting matrix of retransmit, make to receive signal acquisition maximum signal to noise ratio, and compare with the orthogonal space time block coding transmission method of fixed relay gain with the orthogonal space time block coding transmission method based on gain coefficient weighting matrix that traditional CSI is auxiliary, the orthogonal space time block coding transmission method based on the linear weighting matrix of optimum relaying of the present invention's design can obtain more high s/n ratio gain, improve error rate of system performance.

2) the present invention, by use the method for linear weighted function matrix at via node, takes full advantage of the linear process ability at via node place, has improved the detection performance that receives signal.

3) the present invention, by multi-input multi-ouput channel state information is dissolved in optimum linearity weighting matrix, gains thereby take full advantage of multi-antenna diversity, has improved the reliability of input.

Brief description of the drawings

Fig. 1 is the auxiliary orthogonal space time block coding transmission system schematic diagram of existing non-regenerative relaying;

Fig. 2 is that the present invention adopts the orthogonal space time block coding transfer process figure based on the linear weighting matrix of optimum relaying;

Fig. 3 is that the present invention sets the auxiliary orthogonal space time block coding transmission system schematic diagram of non-regenerative relaying;

Fig. 4 is the performance of BER comparison diagram that adopts the inventive method and traditional orthogonal space time block coding method;

Fig. 5 is the error sign ratio performance comparison diagram that adopts the inventive method and traditional orthogonal space time block coding method.

Embodiment

Referring to accompanying drawing, transmission method of the present invention is described in further detail.

With reference to Fig. 2, the orthogonal space time block coding transmission method based on the linear weighting matrix of optimum relaying that the present invention adopts comprises the steps:

Step 1: build the auxiliary orthogonal space time block coding transmission system of non-regenerative relaying.

The auxiliary orthogonal space time block coding transmission system of non-regenerative relaying that the present invention builds, as shown in Figure 3, this system is by a source node S, and a via node R and a destination node D form.Wherein, source node S has two antennas, and via node R has N _rroot antenna, destination node D has an antenna.A transmission cycle of this communication system comprises two time slots, and each time slot comprises two moment.In addition, the present invention suppose any a pair of transmission/reception antennas between channel be flat fading channel, channel fading coefficient remains unchanged in two time slots.And the known source node S of suppose relay node R of the present invention arrives the channel condition information CSI of destination node D to via node R and via node R.

Step 2: source node S is used power P _ssend two unlike signal vectors to via node R in two moment of the first time slot respectively.

In the first moment, send primary signal vector x by source node S ₁give via node R, in the second moment, send the signal vector x after space-time block coding by source node S ₂give via node R, and x ₁and x ₂meet following form:

$x_1=\left[\begin{array}{c}{m}_1\\ m_2\end{array}\right]---\left(1\right)$

$x_2=\left[\begin{array}{c}{-m}_2^{*}\\ {\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">m</figure-callout>}}_1^{*}\end{array}\right]---\left(2\right)$

Wherein, m ₁for first modulation symbol at source node S place, m ₂for second modulation symbol at source node S place, () ^*represent conjugate operation.

Step 3: via node R receives two different pieces of informations from source node S in two moment of the first time slot respectively.

In the first moment, first data y that via node R receives _r1for:

y _r1＝H ₁x ₁+n _r1 (3)

Wherein, H ₁for source node S is to the channel matrix of via node R, dimension is (N _r× 2), N wherein _rfor the antenna number at via node R place, n _r1for first additive white Gaussian noise vector at via node R place, dimension is (N _r× 1), and n _r1average be zero, variance matrix is

wherein for the noise power at via node R place, I is unit matrix;

In the second moment, second data y that via node R receives _r2for:

y _r2＝H ₁x ₂+n _r2 (4)

Wherein, n _r2for second, via node R place additive white Gaussian noise vector, dimension is (N _r× 1), and n _r2average be zero, variance matrix is

Step 4: via node R estimates the channel matrix H of source node S to via node R ₁, and to H ₁carry out singular value decomposition.

Via node R estimates the channel matrix H of source node S to via node R ₁can adopt several different methods of the prior art, for example pilot frequency sequence estimation technique, the training sequence estimation technique, the blind estimation technique, the half-blindness estimation technique.This example adopts the training sequence estimation technique to estimate the channel matrix H of source node S to via node R ₁specific implementation process is: source node S sends known original training sequence to via node R, the decline training sequence after channel matrix effect that via node R receives, then via node R multiplies each other the contrary and known original training sequence of decline training sequence, obtains the channel matrix H of source node S to via node R ₁.

Via node R is the channel matrix H to via node R to source node S ₁carry out singular value decomposition, undertaken by following formula:

$H_1={\mathrm{<figure-callout\; id="1"\; label="U"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">U</figure-callout>}}_1\left[\begin{array}{c}{\widehat{\mathrm{\&Lambda;}}}_1\\ 0\end{array}\right]V_1^H={\mathrm{<figure-callout\; id="1"\; label="U"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">U</figure-callout>}}_1{\mathrm{\&Lambda;}}_1{V}_1^H---\left(5\right)$

Wherein, U ₁it is the channel matrix H that source node S arrives via node R ₁left singular vector,

for the diagonal matrix of (2 × 2) dimension, V ₁it is the channel matrix H that source node S arrives via node R ₁right singular vector, ${\mathrm{\&Lambda;}}_1=\left[\begin{array}{c}{\widehat{\mathrm{\&Lambda;}}}_1\\ 0\end{array}\right]$ For (N _r× 2) matrix of dimension, () ^hoperate for conjugate transpose.

Step 5: via node R obtains the channel matrix h of via node R to destination node D ₂, and to h ₂carry out singular value decomposition.

Via node R can obtain the channel matrix h of via node R to destination node D by several different methods in prior art ₂, for example open loop estimation methods, closed-loop estimation method, feedback information method.This example adopts feedback information method, and specific implementation process is: destination node D first adopts the training sequence estimation technique to obtain the channel matrix h of via node R to destination node D ₂, then destination node D arrives via node R by feedback link the channel matrix h of destination node D ₂feed back to via node R.

Via node R is the channel matrix h to destination node D to relaying node R ₂carry out singular value decomposition, undertaken by following formula:

${\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2=\left[\begin{array}{cc}{\mathrm{\&lambda;}}_{21}& 0\end{array}\right]V_2^H={\mathrm{\&Lambda;}}_2{V}_2^H---\left(6\right)$

Wherein, λ ₂₁the channel matrix h of via node R to destination node D ₂non-zero singular value, V ₂the channel matrix h of via node R to destination node D ₂right singular vector, Λ ₂=[λ ₂₁0] be N _rthe row vector of individual element.

Step 6: via node R builds the optimum linearity weighting square that makes received signal to noise ratio maximum.

Via node R is the channel matrix H to via node R according to source node S ₁left singular vector U ₁with the channel matrix h of via node R to destination node D ₂right singular vector V ₂, build and make the optimum linearity weighting matrix W of received signal to noise ratio maximum be:

$W=\sqrt{\frac{P_r}{E_S{\mathrm{\&lambda;}}_{11}^2+{\mathrm{\&sigma;}}_r^2}}{\left(V_2\right)}_1{\left(U_1\right)}_1^H---\left(7\right)$

Wherein, P _rfor the transmitted power of via node R, E _sfor the energy of source node S transmission symbol, and E _s=P _s/ 2, λ ₁₁for source node S is to the channel matrix H of via node R ₁singular value, for noise power, (V ₂) ₁for via node R is to the channel matrix h of destination node D ₂right singular vector V ₂first row element, (U ₁) ₁for source node S is to the channel matrix H of via node R ₁left singular vector U ₁first row element.

Step 7: in the second time slot, via node R builds two different pieces of informations two moment respectively and sends to destination node D.

In the first moment, first in optimum linearity weighting matrix W and step 3 that via node R constructs step 6 receives data y _r1multiply each other, first sends data x to obtain relaying _r1, and send to destination node D, wherein:

x _r1＝Wy _r1＝WH ₁x ₁+Wn _r1 (8)

In the second moment, second in the optimum linearity weighting matrix W that via node R constructs step 6 and step 3 receives data y _r2multiply each other, obtain second, relaying and send data x _r2, and send to destination node D, wherein:

x _r2＝Wy _r2＝WH ₁x ₂+Wn _r2 (9)

Step 8: destination node D receives two different pieces of informations from via node R in two moment of the second time slot respectively.

In the first moment, destination node D receives first data y _d1for:

${\mathrm{<figure-callout\; id="1"\; label="y\; d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">y</figure-callout>}}_d=h_2{\mathrm{<figure-callout\; id="1"\; label="x\; r"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">x</figure-callout>}}_r+{\mathrm{<figure-callout\; id="1"\; label="n\; d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">n</figure-callout>}}_d={\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1{x}_1+{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{<figure-callout\; id="1"\; label="Wn\; r"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">Wn</figure-callout>}}_r+{\mathrm{<figure-callout\; id="1"\; label="n\; d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">n</figure-callout>}}_d={\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1{x}_1+{\hat{n}}_{d1}---\left(10\right)$

Wherein, n _d1for first additive white Gaussian noise at destination node D place, and its average is zero, and variance is

for first equivalent noise at destination node D place, and its average is zero, and variance is ${\widehat{\mathrm{\&sigma;}}}_{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">d</figure-callout>}}^2={\mathrm{\&sigma;}}_{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">d</figure-callout>}}^2+{\mathrm{\&sigma;}}_r^2{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WW}}^H{h}_2^H;$

In the second moment, destination node D receives second data y _d2for:

${\mathrm{<figure-callout\; id="2"\; label="y\; d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">y</figure-callout>}}_d=h_2{\mathrm{<figure-callout\; id="2"\; label="x\; r"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">x</figure-callout>}}_r+{\mathrm{<figure-callout\; id="2"\; label="n\; d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">n</figure-callout>}}_d={\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1{x}_2+{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{<figure-callout\; id="2"\; label="Wn\; r"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">Wn</figure-callout>}}_r+{\mathrm{<figure-callout\; id="2"\; label="n\; d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">n</figure-callout>}}_d={\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1{x}_2+{\hat{n}}_{d2}---\left(11\right)$

Wherein, n _d2for second additive white Gaussian noise at destination node D place, and its average is zero, and variance is

for second equivalent noise at destination node D place, and its average is zero, and variance is ${\widehat{\mathrm{\&sigma;}}}_{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">d</figure-callout>}}^2={\mathrm{\&sigma;}}_{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">d</figure-callout>}}^2+{{\mathrm{\&sigma;}}_r}^2{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WW}}^H{h}_2^H.$

Step 9: destination node D is according to two data y that receive _d1and y _d2build judgement vector.

Destination node D can build judgement vector, for example ZF method, least mean-square error method, maximum ratio act of union by several different methods in prior art.This example adopts maximum ratio act of union, and specific implementation is carried out according to the following steps:

(9a) destination node D is by first data y receiving _d1with second data y _d2after, be encapsulated as vector form

${\hat{y}}_d=\left[\begin{array}{c}{y}_{d1}\\ {\mathrm{<figure-callout\; id="2"\; label="y\; d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">y</figure-callout>}}_d^2\end{array}\right]\left[\begin{array}{cc}{\left[{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1\right]}_1& {\left[{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1\right]}_2\\ {\left[{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1\right]}_2^{*}& {-\left[{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1\right]}_1^{*}\end{array}\right]\left[\begin{array}{c}{x}_1\\ x_2\end{array}\right]+\left[\begin{array}{c}{\hat{n}}_{d1}\\ {\hat{n}}_{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">d</figure-callout>}2}^{*}\end{array}\right]=\mathrm{Hx}+{\hat{n}}_d---\left(12\right)$

Wherein, [h ₂wH ₁] _ifor vectorial h ₂wH ₁i element, i=1,2, $H=\left[\begin{array}{cc}{\left[{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1\right]}_1& {\left[{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1\right]}_2\\ {\left[{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1\right]}_2^{*}& {-\left[{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1\right]}_1^{*}\end{array}\right]$ Represent compound channel matrix, $x=\left[\begin{array}{c}{m}_1\\ {\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">m</figure-callout>}}_2\end{array}\right]$ For the primary signal vector at source node S place, ${\hat{n}}_d=\left[\begin{array}{c}{\hat{n}}_{d1}\\ {\hat{n}}_{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">d</figure-callout>}2}^{*}\end{array}\right]$ For the recombination noise vector at destination node D place;

(9b) the compound channel matrix H obtaining in step (9a) is carried out conjugate transpose operation by destination node D, and with step (9a) in reception data vector

multiply each other, construct judgement vector

for:

${\tilde{y}}_d=H^H{\hat{y}}_d=\underset{i=1}{\overset{2}{\mathrm{\&Sigma;}}}{\left|{\left[{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HDA0000151125440000032.png"\; state="\{\{state\}\}">h</figure-callout>}}_2{\mathrm{WH}}_1\right]}_i\right|}^{2}x+{\tilde{n}}_d---\left(13\right)$

Wherein, () ^hrepresent conjugate transpose operation, and its average is zero, and variance matrix is

for the equivalent noise variance at destination node D place.

Step 10: destination node D utilizes the judgement vector that step 9 obtains to carry out decoding.

Destination node D can carry out decoding by several different methods in prior art, for example maximum-likelihood decoding method, and Viterbi decoding method, based on the decoding method of minimum mean square error criterion.This example adopts maximum-likelihood decoding method, and specific implementation is: destination node D is by the judgement vector obtaining in step (9b)

carry out Euclidean distance comparison with all signaling points in the planisphere at primary signal vector x place, and find out with judgement vector

the signaling point of Euclidean distance minimum, obtains decode results

for:

$\tilde{x}=\arg \underset{{\tilde{x}}_i\&Element;G}{\min }d({\tilde{y}}_d,{\tilde{x}}_i)---\left(14\right)$

Wherein,

for the some signaling points in planisphere, G is the set of all signaling points in signal constellation (in digital modulation) figure,

represent judgement vector

with the some signaling points in planisphere

euclidean distance between the two,

represent the minimum value of all elements.

Effect of the present invention can further illustrate by following simulation result:

A. simulated conditions: set a relay cooperative system, comprise a source node S, a via node node R and a destination node D.Source node S has two transmitting antennas, the number of antennas N at via node R place _rbe increased to 4 from 2, destination node D has a reception antenna.The modulation system adopting is 4QAM.And the transmitted power of source node S is P _s=1, the transmitted power at via node R place is P _r=1, and suppose relay node R is the same with the noise variance at destination node D place,

suppose that in addition channel is Rayleigh flat fading channel, and the average of channel coefficients is zero, variance is 1.

B. emulation content:

B1) adopt respectively the orthogonal space time block coding transmission method based on the linear weighting matrix of optimum relaying that traditional orthogonal space time block coding transmission method based on gain coefficient weighting matrix and the present invention propose to carry out emulation to the average error bit rate BER of cooperative relay system with respect to average transmission signal to noise ratio, simulation result as shown in Figure 4.

B2) adopt respectively the orthogonal space time block coding transmission method based on the linear weighting matrix of optimum relaying that traditional orthogonal space time block coding transmission method based on gain coefficient weighting matrix and the present invention propose to carry out emulation to the error sign ratio SER of cooperative relay system with respect to average transmission signal to noise ratio, simulation result as shown in Figure 5.

C. simulation result:

As can be seen from Figure 4, the performance of BER curve that adopts the orthogonal space time block coding transmission method based on the linear weighting matrix of optimum relaying of the present invention's proposition to obtain is starkly lower than the performance of BER curve of traditional orthogonal space time block coding transmission method based on gain coefficient weighting matrix, be 4 in via node R antenna number, and bit error rate is 10 ^-3time, adopt the method for the present invention's proposition with respect to conventional method, system can obtain the gain of 12.5dB, shows to adopt the transmission method that the present invention proposes can reduce error rate of system, improves the transmission performance of system.And, while using the method for the present invention's proposition, system bit error rate can reduce along with the increase of via node R place antenna number, and use when conventional method, system bit error rate does not change in the time that via node R place antenna number increases, the diversity gain that shows to adopt method that the present invention proposes can more effectively utilize many, place of via node R antenna to provide, thus the reliability of input improved.

As can be seen from Figure 5, the error sign ratio performance curve that adopts the orthogonal space time block coding transmission method based on the linear weighting matrix of optimum relaying of the present invention's proposition to obtain is starkly lower than the error sign ratio performance curve of traditional orthogonal space time block coding transmission method based on gain coefficient weighting matrix, be 4 in via node R antenna number, and error sign ratio is 10 ^-3time, adopt the method for the present invention's proposition with respect to conventional method, system can obtain the gain of 12.5dB, shows to adopt the transmission method that the present invention proposes can reduce system error sign ratio, improves the transmission performance of system.And, while using the method for the present invention's proposition, system error sign ratio can reduce along with the increase of via node R place antenna number, and use when conventional method, system error sign ratio does not change in the time that via node R place antenna number increases, the diversity gain that shows to adopt method that the present invention proposes can more effectively utilize many, place of via node R antenna to provide, thus the reliability of input improved.

In sum, the present invention compares with traditional orthogonal space time block coding transmission method based on gain coefficient weighting matrix, has reduced bit error rate and the error rate of system, thereby has improved the transmission performance of system.

Claims (2)

1. the orthogonal space time block coding transmission method based on the linear weighting matrix of optimum relaying, comprises the steps:

(1) set a source node S, a via node R and a destination node D, to build the auxiliary orthogonal space time block coding transmission system of non-regenerative relaying, and be divided into two time slots by a transmission cycle of this system, and each time slot comprises two moment;

(2) source node S transmitted signal in the first time slot, in the first moment, source node S sends primary signal vector x ₁give via node R, in the second moment, source node S sends the signal vector x after space-time block coding ₂give via node R, and x ₁and x ₂meet following form:

$x_1=\left[\begin{array}{c}{m}_1\\ m_2\end{array}\right]$

$x_2=\left[\begin{array}{c}-m_2^{*}\\ m_1^{*}\end{array}\right]$

Wherein, m ₁for first modulation symbol at source node S place, m ₂for second modulation symbol at source node S place, () ^*represent conjugate operation;

(3) via node R receives data in the first time slot, and in the first moment, first data that via node R receives are y _r1, in the second moment, second data that via node R receives are y _r2;

(4) via node R uses the training sequence estimation technique to obtain the channel matrix H of source node S to via node R ₁, and it is carried out to singular value decomposition:

$H_1=U_1\left[\begin{array}{c}{\widehat{\mathrm{\&Lambda;}}}_1\\ 0\end{array}\right]{V_1}^H=U_1{\mathrm{\&Lambda;}}_1{V_1}^H$

Wherein, U ₁it is the channel matrix H that source node S arrives via node R ₁left singular vector,

for the diagonal matrix of (2 × 2) dimension, V ₁it is the channel matrix H that source node S arrives via node R ₁right singular vector, ${\mathrm{\&Lambda;}}_1=\left[\begin{array}{c}{\widehat{\mathrm{\&Lambda;}}}_1\\ 0\end{array}\right]$ For (N _r× 2) matrix of dimension, N _rfor the antenna number at via node R place, () ^hoperate for conjugate transpose;

(5) via node R obtains the channel matrix h of via node R to destination node D by feedback ₂, and it is carried out to singular value decomposition:

$h_2=\left[\begin{array}{cc}{\mathrm{\&lambda;}}_{21}& 0\end{array}\right]V_2^H={\mathrm{\&Lambda;}}_2{V}_2^H$

Wherein, λ ₂₁the channel matrix h of via node R to destination node D ₂non-zero singular value, V ₂the channel matrix h of via node R to destination node D ₂right singular vector, Λ ₂=[λ _{2 1}0] be N _rthe row vector of individual element;

(6) via node R arrives the channel matrix H of via node R according to source node S ₁left singular vector U ₁with the channel matrix h of via node R to destination node D ₂right singular vector V ₂, build and make the optimum linearity weighting matrix W of received signal to noise ratio maximum be:

$W=\sqrt{\frac{P_r}{E_S{\mathrm{\&lambda;}}_{11}^2+{\mathrm{\&sigma;}}_r^2}}{\left(V_2\right)}_1{\left(U_1\right)}_1^H$

Wherein, P _rfor the transmitted power of via node R, E _sfor the energy of source node S transmission symbol, λ ₁₁for source node S is to the channel matrix H of via node R ₁singular value,

for noise power, (V ₂) ₁for via node R is to the channel matrix h of destination node D ₂right singular vector V ₂first row element, (U ₁) ₁for source node S is to the channel matrix H of via node R ₁left singular vector U ₁first row element;

(7) in the second time slot, via node R sends data, and in the first moment, first in the optimum linearity weighting matrix W that via node R constructs step (6) and step (3) receives data y _r1multiply each other, obtain first transmission signal vector of relaying x _r1, and send to destination node D; In the second moment, second in the optimum linearity weighting matrix W that via node R constructs step (6) and step (3) receives data y _r2multiply each other, obtain second transmission signal vector x of relaying _r2, and send to destination node D;

(8) destination node D receives data in the second time slot, and in the first moment, it is y that destination node D receives first data _d1, in the second moment, it is y that destination node D receives second data _d2, and according to receive two data y _d1and y _d2structure judgement vector, to carry out decoding;

Described destination node D is according to two data y that receive _d1and y _d2structure judgement vector, comprises the steps:

(8a) destination node D is by first data y receiving _d1with second data y _d2be encapsulated as vector form

obtain:

${\hat{y}}_d=\left[\begin{array}{c}{y}_{d1}\\ y_{d2}^{*}\end{array}\right]=\left[\begin{array}{cc}{\left[h_2{\mathrm{WH}}_1\right]}_1& {\left[h_2{\mathrm{WH}}_1\right]}_2\\ {\left[h_2{\mathrm{WH}}_1\right]}_2^{*}& -{\left[h_2{\mathrm{WH}}_1\right]}_1^{*}\end{array}\right]\left[\begin{array}{c}{m}_1\\ m_2\end{array}\right]+\left[\begin{array}{c}{\hat{n}}_{d1}\\ {\hat{n}}_{d2}^{*}\end{array}\right]=\mathrm{Hx}+{\hat{n}}_d$

Wherein, [h ₂wH ₁] _i, i=1,2 is vectorial h ₂wH ₁i element, for first equivalent noise at destination node D place,

for second equivalent noise at destination node D place, $H=\left[\begin{array}{cc}{\left[h_2{\mathrm{WH}}_1\right]}_1& {\left[h_2{\mathrm{WH}}_1\right]}_2\\ {\left[h_2{\mathrm{WH}}_1\right]}_2^{*}& -{\left[h_2{\mathrm{WH}}_1\right]}_1^{*}\end{array}\right]$ Represent compound channel matrix, $x=\left[\begin{array}{c}{m}_1\\ m_2\end{array}\right]$ For the primary signal vector at source node S place, ${\hat{n}}_d=\left[\begin{array}{c}{\hat{n}}_{d1}\\ {\hat{n}}_{d2}^{*}\end{array}\right]$ For the recombination noise vector at destination node D place;

(8b) destination node D structure judgement vector

for:

${\tilde{y}}_d=H^H{\hat{y}}_d=\underset{i=1}{\overset{2}{\mathrm{\&Sigma;}}}{\left|{[h}_2{\mathrm{WH}}_1{]}_i\right|}^{2}x+{\tilde{n}}_d$

Wherein, () ^hfor conjugate transpose operation,

2. method according to claim 1, in wherein said step (8), destination node D carries out decoding according to judgement vector, is that destination node D is by the judgement vector obtaining in step (8b)

carry out Euclidean distance comparison with all signaling points in the planisphere at primary signal vector x place, and find out with judgement vector

the signaling point of Euclidean distance minimum, obtains decode results

for:

$\tilde{x}=\arg \underset{{\tilde{x}}_i\&Element;G}{\min }d({\tilde{y}}_d,{\tilde{x}}_i)$

Wherein,

represent the some signaling points in planisphere, G represents the set of all signaling points in signal constellation (in digital modulation) figure,

represent judgement vector with the some signaling points in planisphere

euclidean distance between the two,

represent the minimum value of all elements.

Images