CN101232356A - Precoding method, system and apparatus in MIMO system - Google Patents

Abstract

The invention discloses a pre-coding method, a system and a device in a MIMO system. The method, the system and the device which are related to by the invention can be applied in the orthogonal space-time coding circumstance. The inventive method adopts a non-unitary matrix for pre-coding, and the non-unitary matrix is obtained by judging the obtained optimal pre-coding vector construction. The inventive system includes a transmitter and a receiver, wherein, the transmitter includes a code list, a pre-coding matrix construction unit and a pre-coding unit; the receiver includes a channel estimation unit, the code list and a judgement unit. The simulation and the analysis show that, compared with the prior art, the usage of the inventive proposal can lower the bit error rate and reduce the complexity of decoding simultaneously in the situation of the same signal-to-noise ratio.

Description

Method for precoding in the mimo system, system and device

Technical field

The present invention relates to wireless communication technology, particularly relate to the precoding technique in a kind of mimo system.

Background technology

Information theory studies have shown that adopts the communication system of many antennas can obtain higher channel capacity, at channel is that independent Rayleigh fading and receiver are known under the hypothesis of channel condition information, all adopt multiple-input, multiple-output (the Multiple Input MultipleOutput of many antennas at transmitting terminal and receiving terminal, MIMO) the theoretical channel capacity of system increases with smaller value approximately linear in number of transmit antennas and the reception antenna number, and the mode that mimo system improves capacity is called space division multiplexing.Mimo system and OFDM (Orthogonal Frequency Division Multiplexing, OFDM) the MIMO-OFDM system that combines of technology, can effectively resist the multipath fading of wireless channel, be acknowledged as the most competitive technology of the 4th third-generation mobile communication.

MIMO can utilize the decline of a plurality of independent channels gain antagonism wireless channels, is referred to as diversity, comprises three kinds of space diversity, time diversity and frequency diversities.Wherein, vertical demixing time space (Vertical Bell Laboratory Layer Space-Time Codes, V-BLAST) structure and orthogonal space time packet (Orthogonal Space-Time Block Codes, OSTBC) combine space division multiplexing and time diversity, these two kinds of coded systems all belong to space time coding scheme.

Usually, the transmitting terminal of mimo system is not known channel condition information, yet, if transmitting terminal can be known channel condition information, then we can necessarily handle transmitting at transmitting terminal, reduce and disturb, improve average received signal to noise ratio (Signal to Noise Ratio, SNR), thereby reduce average error sign ratio, we claim the precoding that is treated to of this transmitting terminal.The required channel condition information of transmitting terminal can be instantaneous channel estimating, also can be the single order or the second-order statistics information of channel estimating, this depends on the time fading characteristic of channel, for example, for the rapid fading system, we generally adopt the statistical property of channel estimating to reflect channel information; And for the slow fading system, we generally adopt instantaneous channel estimating.

A symmetrical time division duplex (Time Division Duplex, TDD) in the system, the channel information of up-downgoing channel can be shared mutually, in this case, does not need system configuration is done any change, base station and travelling carriage can both adopt precoding.Yet for asymmetric tdd systems or Frequency Division Duplexing (FDD) (Frequency Division Duplex, FDD) system, situation is then different fully, with this type of system down link precoding is example, because the asymmetry of up-downgoing channel, receiving terminal need be passed to transmitting terminal so that it knows down channel with channel information by a feedback channel, we claim that the mimo system of this employing feedback mechanism is closed-loop MIMO (Closed-loop MIMO, CL-MIMO) system, corresponding, not having the mimo system of feedback is open loop MIMO (Open MIMO, OP-MIMO) system.Closed-loop MIMO system can better be brought into play the advantage of MIMO by feedback optimized systematic function, therefore becomes one of focus of recent MIMO research.

Yet, channel information increases with the long-pending of number of transmit antennas, reception antenna number and number of users, count for the linear growth with number of transmit antennas or reception antenna for the mimo channel capacity, feeding back completely, channel information (ideal feedback) is difficult to realize in real system.The limited feedback precoding scheme can address the above problem.The closed-loop MIMO system that adopts limited feedback precoding is referred to as code book in advance with one group of pre-coding matrix, is stored in transmitting-receiving two-end.According to channel estimating and certain decision rule, select optimum pre-coding matrix at receiving terminal system from code book, its index is fed back to transmitting terminal, transmitting terminal obtains pre-coding matrix according to this index again.With the code book that comprises N pre-coding matrix is example, and the limited feedback precoding scheme only need be fed back log ₂The N bit just can obtain pre-coding matrix, has significantly reduced feedback information volume.We can say that limited feedback precoding is the quantification to desirable pre-coding matrix space.

Present most precoding research document and IEEE802.16e consensus standard draft all adopt above-mentioned this pre-coding scheme, the IEEE802.16e physical layer with its respectively with three kinds of space time coding schemes, be beam forming (beamforming, BF), vertical demixing time space and orthogonal space time packet combine.

The structure of a limited feedback precoding MIMO-OFDM system and workflow are as shown in Figure 1.

At transmitting terminal, at first the data that will send are produced modulated symbol by modulation, mapping (Space Time Coding) when then modulated symbol being carried out sky, the signal that mapping obtains during to sky carries out precoding, after invert fast fourier transformation, D A conversion, uppermixing is launched by sky bundle of lines signal at last.Wherein, in the precoding stage, transmitting terminal is received the feedback information that receiving terminal sends by feedback channel, and which pre-coding matrix this information notification transmitting terminal should select carry out precoding.After transmitting terminal is received this information, in code book set, choose this pre-coding matrix, and the signal that mapping obtains when adopting this matrix to sky carries out precoding.

At receiving terminal, after antenna receives signal, on the one hand, signal is descended mixing, A the D conversion, fast fourier transform, the signal after the conversion is as one of input signal of decoder, and receiving terminal is according to channel estimating and optimum decision criterion on the other hand, from code book, select optimum pre-coding matrix, its sequence number is fed back to transmitting terminal by feedback channel, simultaneously, optimum pre-coding matrix of last time and channel estimating are sent into decoder be used for decoding.Its concrete judging process is as follows: receiving device carries out channel estimating according to the signal that receives, the channel condition information of utilizing channel estimating to obtain, carry out optimum decision in conjunction with each pre-coding matrix in the code book set, and court verdict is sent to transmitting terminal by feedback channel.

The key of limited feedback precoding is code book and optimum decision criterion, and the two will directly influence the bit error rate performance of system.If the optimum pre-coding matrix that obtains through judgement is F, the IEEE802.16e draft agreement thinks that according to the precoding theory, precoding does not increase total emission power, promptly satisfies

$\underset{s\∈S^{M_s}}{\max }\frac{{\left|\right|\mathrm{Fs}\left|\right|}_2}{{\left|\right|s\left|\right|}_2}\≤1---\left(1\right)$

Wherein, s represents one group of vector that transmits, and S represents the modulation symbol collection, ‖ ‖ ₂Vector 2 norms are got in expression.Because unitary matrice has the characteristic that does not change the vector norm, therefore, normal unitary matrice (when number of transmit antennas equals the data fluxion) or the para-unitary matrix (number of transmit antennas greater than fluxion time) of adopting is as pre-coding matrix in precoding.

The IEEE802.16e draft agreement also is the unitary matrice precoding of adopting.Therefore, the pre-coding matrix F of draft agreement regulation is unitary matrice (when number of transmit antennas equals the data fluxion) or para-unitary matrix (number of transmit antennas greater than fluxion time).

In addition, and code book is classified according to different transmit antennas number, different pieces of information fluxion and the configuration of different number of bits of feedback, the scheme of classification of code book has also been proposed in the IEEE802.16e draft agreement, defined according to code book under different transmit antennas number, different pieces of information fluxion and the configuration of different number of bits of feedback, the available code book that its draft provides sees table 1 for details, and "-" is illustrated in and do not stipulate this code book in this agreement.

Table 1IEEE802.16e agreement code book

Therefore, after receiving terminal received signal, according to number of transmit antennas, data fluxion and number of bits of feedback, the wherein class shown in the option table 1 in the code book was carried out optimum decision.

With 21 Alamouti orthogonal space time group codings of receiving of closed loop is example, if its number of bits of feedback is 3bit, then optimum pre-coding matrix F selects from (2,2,3) code book set of table 1; If its number of bits of feedback is 6bit, then F selects from (2,2,6) code book set of table 1.

The purpose of optimum decision is to make minimum bit-error rate or maximum capacity, and is specifically different according to decoding algorithm, and it embodies minimum bit-error rate optimum decision criterion formula and be

$F=\underset{F_i\∈\mathrm{\Φ}}{\arg \max }{\left|\right|{\mathrm{HF}}_i\left|\right|}_F---\left(2\right)$

Wherein, Ф represents code book set, ‖ ‖ _FMatrix F robenius norm is got in expression.Court verdict shows that F is optimum pre-coding matrix, then should feed back to transmitting terminal by the sequence number i of optimum pre-coding matrix in code book.

Summary of the invention

In the background technology part, the orthogonal space time group coding in the above-mentioned example is a kind of of quadrature Space Time Coding, and the orthogonal property of quadrature Space Time Coding has the advantages that not change the channel distribution characteristic.Simultaneously, unitary matrice also has the characteristics that do not change the channel distribution characteristic.In fact, if wish not change the channel distribution characteristic when coding, the coding that only need once not change the channel distribution characteristic gets final product, and does not need to carry out above such coding twice.Therefore, if we have adopted the quadrature Space Time Coding in the MIMO-OFDM system, then can in precoding, adopt non-unitary matrice.

Simultaneously, in wireless transmission, always wish under same signal to noise ratio, to obtain better bit error rate performance, promptly obtain big as far as possible snr gain.The purpose of the specific embodiment of the present invention is that a kind of mthods, systems and devices that utilize non-unitary matrice to carry out precoding in aforementioned system will be provided, and not change under the condition of channel distribution characteristic in assurance, obtains better bit error rate performance.

The specific embodiment of the present invention provides the method for precoding in a kind of mimo system, and this method adopts the quadrature Space Time Coding, and adopts two above data fluxions, and comprises:

In the beam shaping code book, choose optimum precoding vectors;

Adopt described optimum precoding vectors structure pre-coding matrix;

Adopt described pre-coding matrix to carry out precoding.

The specific embodiment of the present invention provides a kind of mimo system that adopts limited feedback precoding again, and this system comprises transmitter and receiver.This system adopts the quadrature Space Time Coding, and adopts two above data fluxions; And,

Described receiver comprises:

Channel estimating unit is used to carry out channel estimating, and channel estimation results is offered decision unit;

Code book list storage unit is used for providing pre-coding matrix or precoding vectors to decision unit;

Decision unit, the channel estimation results that provides according to described channel estimating unit and the decision rule of setting are provided, in the beam shaping code book of described code book list storage unit, choose optimum precoding vectors, and feedback information that should the optimum precoding vectors is given described transmitter;

Described transmitter comprises:

Code book list storage unit is used for providing pre-coding matrix or precoding vectors to the pre-coding matrix structural unit;

The pre-coding matrix structural unit is used for the optimum precoding vectors information according to described receiver decision unit feedback, chooses corresponding precoding vectors in the code book list storage unit of described transmitter, the structure pre-coding matrix;

Precoding unit is used for the pre-coding matrix according to described pre-coding matrix structural unit structure, carries out precoding.

The specific embodiment of the present invention provides a kind of receiver that is used in the mimo system again, and described receiver comprises:

Channel estimating unit is used to carry out channel estimating, and channel estimation results is offered decision unit;

Code book list storage unit is used for providing pre-coding matrix or precoding vectors to decision unit;

Decision unit, the channel estimation results that provides according to described channel estimating unit and the decision rule of setting are provided, in the beam shaping code book of described code book list storage unit, choose optimum precoding vectors, and feedback information that should the optimum precoding vectors is given described transmitter;

The specific embodiment of the present invention provides a kind of transmitter that is used in the mimo system again, it is characterized in that, described transmitter comprises:

Code book list storage unit is used for providing pre-coding matrix or precoding vectors to the pre-coding matrix structural unit;

The pre-coding matrix structural unit is used for the optimum precoding vectors information according to described receiver decision unit feedback, chooses corresponding precoding vectors in the code book list storage unit of described transmitter, the structure pre-coding matrix;

Precoding unit is used for the pre-coding matrix according to described pre-coding matrix structural unit structure, carries out precoding.

The specific embodiment of the invention provides a kind of method, system, transmitter and receiver that utilizes beam shaping code book structure pre-coding matrix.The pre-coding matrix of beam shaping code book structure is a kind of non-unitary matrice.Therefore, adopt method, system, transmitter and receiver that the specific embodiment of the invention provided, can realize utilizing non-unitary matrice to carry out precoding.After carrying out precoding with the described method of embodiment of the present invention, can obtain, and can know that by analyzing its decoding complexity is lower than the better bit error rate performance of unitary matrice pre-coding scheme.

Description of drawings

Fig. 1 is a limited feedback precoding MIMO-OFDM system schematic of the prior art;

Fig. 2 is under maximum doppler frequency is 40 low speed mobile environment, and signal to noise ratio and bit error rate concerns comparison diagram under I type ideal feedback and the II type ideal feedback;

Fig. 3 is under maximum doppler frequency is 40 low speed mobile environment, respectively to the specific embodiment of the invention provided (2,1,3) Limited Feedback is with (2,2 of available technology adopting, 3) carry out emulation under the Limited Feedback situation, the signal to noise ratio and the bit error rate that obtain concern comparison diagram;

Fig. 4 is under maximum doppler frequency is 40 low speed mobile environment, respectively to the specific embodiment of the invention provided (2,1,6) Limited Feedback is with (2,2 of available technology adopting, 6) carry out emulation under the Limited Feedback situation, the signal to noise ratio and the bit error rate that obtain concern comparison diagram;

Fig. 5 is under independent 12 footpath TU channels, when translational speed is 3km/h, respectively to the specific embodiment of the invention provided (2,1,3) Limited Feedback is with (2 of available technology adopting, 2,3) carry out emulation under the Limited Feedback situation, the signal to noise ratio and the bit error rate that obtain concern comparison diagram;

Fig. 6 is under independent 12 footpath TU channels, when translational speed is 3km/h, respectively to the specific embodiment of the invention provided (2,1,6) Limited Feedback is with (2 of available technology adopting, 2,6) carry out emulation under the Limited Feedback situation, the signal to noise ratio and the bit error rate that obtain concern comparison diagram;

Fig. 7 is the mimo system and the schematic representation of apparatus of the specific embodiment of the invention three described employing limited feedback precodings.

Embodiment

For making technical scheme of the present invention clearer, below provide the specific embodiment of the present invention.

Embodiment one:

Embodiment one has provided in the MIMO-OFDM system, when adopting the quadrature Space Time Coding, how to utilize non-unitary matrice to carry out precoding.Specific as follows:

Receiving terminal is received after the signal, carries out channel estimating, obtains M _r* M _fChannel matrix H.H can obtain by any known channel estimation methods, repeats no more herein; According to number of transmit antennas and number of bits of feedback, the pre-coding matrix in the corresponding beam shaping code book set is carried out optimum decision, can obtain optimum emission vector v, the following formula of judgement foundation:

$v=\underset{v_i\∈\mathrm{\ζ}}{\arg \max }{\left|\right|H{v}_i\left|\right|}_2---\left(3\right)$

Wherein, v _iBe precoding vectors in the beam forming code book, i.e. the data fluxion code book that equals 1.The beam forming code book can be the code book that provides in present any document, reference of the present invention be beam forming code book in the IEEE802.16e draft agreement.

Obtain just can constructing optimum pre-coding matrix after the optimum emission vector

F＝[v…v](4)

Formula (4) expression constitutes optimum pre-coding matrix F side by side by n optimum precoding vectors v, and the n value is identical with the data fluxion.Transmitting terminal utilizes F to carry out precoding and gets final product.

Embodiment two:

Below the Alamouti orthogonal space time group codings received with 21 of closed loops be example, specify method for precoding involved in the present invention.In order to compare the performance of the specific embodiment of the present invention and prior art hereinafter, suppose earlier at this: the optimum pre-coding matrix that judgement produces according to prior art is F _I, and the optimum pre-coding matrix that described method judgement produces according to the specific embodiment of the invention is F _II

In the MIMO-OFDM system that adopts 21 receipts of closed loop Alamouti orthogonal space time group coding, number of transmit antennas is 2, because and this embodiment is to be the optimum pre-coding matrix of base configuration with optimum precoding vectors in the beam shaping code book, therefore when number of bits of feedback is 3bit, (2,1,3) select a vector as optimum precoding vectors v in the code book set, when number of bits of feedback is 6bit, (2,1,6) select a vector as optimum precoding vectors v in the code book set.In the embodiment of the present invention, formula (3) is followed in the decision rule of v.

At this code book is described further.When number of bits of feedback was 3bit, this represented that 8 different precoding vectors (or matrix) are arranged in the corresponding code book.For example in (2,1,3) code book 8 different precoding vectors are arranged.By that analogy, in (2,1,6) code book 64 different precoding vectors are arranged, v in the formula (3) _iSubscript i represent the sequence number of this vector in code book.

After judgement produced optimum emission vector v, with court verdict, the sequence number of promptly optimum emission vector v fed back to transmitting terminal to receiving terminal by feedback channel.

Transmitting terminal is constructed optimum pre-coding matrix F according to the described principle of formula (4) after receiving the court verdict of receiving terminal feedback _II, and utilize this matrix to carry out precoding.

Below the MIMO-OFDM system of the Alamouti orthogonal space time group codings received with 21 of embodiment two described employing closed loops be example, the beneficial effect of the described method for precoding of the specific embodiment of the invention is described.

Because the theoretical performance analysis of Limited Feedback is very difficult, provides its error rate lower limit here, just the performance evaluation under the ideal feedback.We claim that the pre-coding scheme in the prior art IEEE802.16e agreement is the precoding of I type closed loop ideal feedback, and its pre-coding matrix is F _I,, take from (2,2,3) or (2,2,6) code book according to the number of bits of feedback difference, F _I=[v ₁v ₂], v ₁, v ₂Be 2 * 1 vector, and mutually orthogonal; Claim that the described pre-coding scheme of the specific embodiment of the invention is the precoding of II type closed loop ideal feedback, its pre-coding matrix is F _II, F _II=[v v], v take from (2,1,3) or (2,1,6) code book according to the number of bits of feedback difference, are the optimum emission vector that obtains through optimum decision, also are 2 * 1 vectors.

2 the 1 Alamouti orthogonal space time group coding codeword structures of receiving are

$S=\left[\begin{array}{cc}{s}_1& {-s}_2^{*}\\ s_2& s_1^{*}\end{array}\right]---\left(5\right)$

Wherein, first column vector is moment t ₁From the signal that 2 transmit antennas are launched, the secondary series vector is moment t ₂The signal of launching from 2 transmit antennas, promptly ${s_t}_1={\left[\begin{array}{cc}{s}_1& s_2\end{array}\right]}^T,$ ${s_t}_2={\left[\begin{array}{cc}{-s}_2^{*}& s_1^{*}\end{array}\right]}^T,$ () T represents to ask transposition.According to Alamouti orthogonal space time group coding hypothesis, t ₁, t ₂Constantly experience identical channel fading, then under the situation of limited feedback precoding, the signal that receiving terminal is received can be expressed as

${y_t}_i=\sqrt{\frac{1}{2}}{{\mathrm{HFs}}_t}_i+{n_t}_i,i=\mathrm{1,2}---\left(6\right)$

Wherein, H is 1 * 2 actual channel matrix, H=[h ₁h ₂], h ₁, h ₂Represent the channel fading coefficient between transmitting antenna 1,2 and the reception antenna respectively,

The expression white Gaussian noise.In the embodiments of the present invention, suppose that fading coefficients is independent, obey average and be 0, variance is 1 multiple Gaussian Profile, promptly H is a rayleigh fading channel.

According to precoding and Alamouti orthogonal space time group coding theory, can be following form with the equivalence of (6) formula

$y=\left[\begin{array}{c}{y}_{t_1}\\ y_{t_2}^{*}\end{array}\right]=\sqrt{\frac{1}{2}}{H}_{\mathrm{eff}}s+n=\sqrt{\frac{1}{2}}\·\left[\begin{array}{cc}{\tilde{h}}_1& {\tilde{h}}_2\\ {\tilde{h}}_2^{*}& {\tilde{h}}_1^{*}\end{array}\right]\·\left[\begin{array}{c}{s}_1\\ s_2\end{array}\right]+\left[\begin{array}{c}{n}_{t_1}\\ n_{t_2}^{*}\end{array}\right]---\left(7\right)$

Wherein, To 2 * 2 equivalent channel matrix H in the formula (7) _Eff(Singular Value Decomposition SVD) decomposes, and gets H to do singular value _Eff=U ∑ V ^H, wherein, U=1, V are 2 * 2 unitary matrice, ∑=[λ 0], and λ is equivalent channel matrix H _EffNon-zero singular value (λ is a real number), according to the Singular Value Decomposition Using theory, have $\mathrm{\λ}=\sqrt{{\left|{\tilde{h}}_1\right|}^2+{\left|{\tilde{h}}_2\right|}^2}.$

Under I type closed loop ideal feedback precoding situation, F _I=V=[v ₁v ₂], HF _I=[λ 0], its decoding back result is

${\hat{S}}_I=H_{\mathrm{eff}}^{H}y$

$=\left[\begin{array}{cc}\mathrm{\λ}& 0\\ 0& \mathrm{\λ}\end{array}\right]\·\sqrt{\frac{1}{2}}\·\left[\begin{array}{cc}\mathrm{\λ}& 0\\ 0& \mathrm{\λ}\end{array}\right]\·{\left[\begin{array}{cc}{s}_1& s_2\end{array}\right]}^T+\left[\begin{array}{cc}\mathrm{\λ}& 0\\ 0& \mathrm{\λ}\end{array}\right]\·n---\left(8\right)$

$=\sqrt{\frac{1}{2}}\·\left[\begin{array}{cc}{\mathrm{\λ}}^2& 0\\ 0& {\mathrm{\λ}}^2\end{array}\right]\·{\left[\begin{array}{cc}{s}_1& s_2\end{array}\right]}^T+\left[\begin{array}{c}\mathrm{\λ}\·n_{t_1}\\ \mathrm{\λ}\·n_{t_2}^{*}\end{array}\right]$

Wherein, () ^HThe matrix transpose conjugate matrices are asked in expression, and we claim in (8) formula

First row and second behavior decipher branch road, the signal of every branch road is sent into the demodulator demodulation after decoding, bit error rate performance then depends on the demodulation signal to noise ratio.According to (8) formula, and hypothesis emission symbol power normalizing, i.e. ε _s=1 and noise power be 1, promptly ${\mathrm{\σ}}_n^2=1$ As can be known, the every branch road demodulation signal to noise ratio in decoding back is

$P_I=\frac{P_s}{P_n}=\frac{{({\mathrm{\λ}}^2/\sqrt{2})}^2\·{\mathrm{\ϵ}}_s}{{\mathrm{\λ}}^2\·{\mathrm{\σ}}_n^2}=\frac{{\mathrm{\λ}}^2}{2}---\left(9\right)$

Under II type closed loop ideal feedback precoding situation, F _II=V=[v v], HF _II=[λ λ], its decoding back result is

${\hat{S}}_I=H_{\mathrm{eff}}^{H}y$

$=\left[\begin{array}{cc}\mathrm{\λ}& \mathrm{\λ}\\ \mathrm{\λ}& -\mathrm{\λ}\end{array}\right]\·\sqrt{\frac{1}{2}}\·\left[\begin{array}{cc}\mathrm{\λ}& \mathrm{\λ}\\ \mathrm{\λ}& -\mathrm{\λ}\end{array}\right]\·{\left[\begin{array}{cc}{s}_1& s_2\end{array}\right]}^T+\left[\begin{array}{cc}\mathrm{\λ}& \mathrm{\λ}\\ \mathrm{\λ}& -\mathrm{\λ}\end{array}\right]\·n---\left(10\right)$

$=\sqrt{\frac{1}{2}}\·\left[\begin{array}{cc}2{\mathrm{\λ}}^2& 0\\ 0& 2{\mathrm{\λ}}^2\end{array}\right]\·{\left[\begin{array}{cc}{s}_1& s_2\end{array}\right]}^T+\left[\begin{array}{c}\mathrm{\λ}\·(n_{t_1}+n_{t_2}^{*})\\ \mathrm{\λ}\·(n_{t_1}-n_{t_2}^{*})\end{array}\right]$

The every branch road signal to noise ratio in decoding back is

$P_{\mathrm{II}}=\frac{P_s}{P_n}=\frac{{(2{\mathrm{\λ}}^2/\sqrt{2})}^2\·{\mathrm{\ϵ}}_s}{2{\mathrm{\λ}}^2\·{\mathrm{\σ}}_n^2}={\mathrm{\λ}}^2---\left(11\right)$

As seen, the demodulation signal to noise ratio that adopts the precoding of II type closed loop ideal feedback to double than I type, promptly systematic function improves 3dB.Because traditional method for precoding and described method for precoding reality of the specific embodiment of the invention is respectively the quantification of I type and the precoding of II type closed loop ideal feedback, therefore, if quantization error is identical, two kinds of limited feedback precoding performance differences also are 3dB.That is to say that the precoding of II type closed loop ideal feedback has better bit error rate performance than I type.

Fig. 2, Fig. 3, Fig. 4 have provided the bit error rate performance contrast that maximum doppler frequency is the specific embodiment of the invention described method and traditional method for precoding under 40 the low speed mobile environment, Fig. 2 is the performance comparison under desirable feedback states, when Fig. 3 is the 3bit feedback, the performance comparison of scheme provided by the present invention and prior art, when Fig. 4 is the 6bit feedback, the performance comparison of scheme provided by the present invention and prior art.Can see that the described method of the specific embodiment of the invention has about 3dB during high s/n ratio under the rayleigh fading channel of separate single footpath with respect to traditional method for precoding and gains.In order further to verify the performance of the present invention program under independent multipath channel, Fig. 5, Fig. 6 have provided under independent 12 footpath TU channels, the bit error rate performance of the present invention program and traditional scheme contrast when translational speed is 3km/h, system's carrier frequency 3.5GHz wherein, bandwidth 10MHz, counting of Fourier transform and Cyclic Prefix is respectively 1024 and 128.Wherein, when Fig. 5 is the 3bit feedback, the performance comparison of scheme provided by the present invention and prior art, when Fig. 6 is the 6bit feedback, the performance comparison of scheme provided by the present invention and prior art.From Fig. 5, Fig. 6 as can be seen, under independent multipath channel, the present invention program is still effective, still can obtain about about 3dB with respect to traditional scheme under high s/n ratio and gain.

Meanwhile, the precoding of II type closed loop ideal feedback is not to be cost to improve total emission power to the raising of signal to noise ratio.According to the structure of Alamouti quadrature Space Time Coding, t ₁Total emission power constantly is

${\left|\right|\frac{1}{\sqrt{2}}{\mathrm{Fs}}_{t_1}\left|\right|}_2^2=\frac{{\left|\right|{\mathrm{vs}}_1+{\mathrm{vs}}_2\left|\right|}_2^2}{2}$

$\≤\frac{{\left|\right|{\mathrm{vs}}_1\left|\right|}_2^2+{\left|\right|{\mathrm{vs}}_2\left|\right|}_2^2}{2}---\left(11\right)$

$=\frac{{\left|\right|v\left|\right|}_2^2({\left|s_1\right|}^2+{\left|s_2\right|}^2)}{2}=1$

Wherein, S _tExpression t ₁The vector that transmits of moment Alamouti orthogonal space time group coding, t ₂Constantly in like manner can demonstrate,prove.Hence one can see that, adopts the precoding of II type closed loop ideal feedback not improve total emission power with respect to the precoding of I type closed loop ideal feedback.

In addition, adopt the described method for precoding of the specific embodiment of the invention can also reduce decoding complexity when reducing error rate of system, this is to cause owing to the pre-coding matrix column vector is identical.When table 2 has provided two data flow, the contrast of the described method for precoding of the traditional method for precoding and the specific embodiment of the invention on decoding complexity.

The contrast of table 2Aloumouti orthogonal space time group coding decoding complexity

For realizing said method, provide the embodiment of corresponding wireless communication system and device below.

Below in conjunction with accompanying drawing 7 the specific embodiment of the present invention three are described:

The specific embodiment of the present invention three provides a kind of mimo system that adopts limited feedback precoding, and this system comprises transmitter 710 and receiver 720.System adopts the quadrature Space Time Coding, and adopts two above data fluxions, and described receiver 720 comprises:

Channel estimating unit 721 is used to carry out channel estimating, and channel estimation results is offered decision unit 723; In addition, channel estimating unit also offers decoder with channel estimation results, but channel estimating unit effect in this respect is not directly related with technical scheme involved in the present invention, therefore should not be considered as the part of the described technical scheme of this embodiment yet;

Code book list storage unit 722 is used for providing pre-coding matrix or precoding vectors to decision unit 723; Described precoding vectors in fact also is a kind of pre-coding matrix, and in embodiment of the present invention, precoding vectors can be selected from the beam shaping code book, is used for constructing pre-coding matrix;

Decision unit 723, the channel estimation results that provides according to described channel estimating unit 721 and the decision rule of setting are provided, in the beam shaping code book of described code book list storage unit 722, choose optimum precoding vectors, and feedback information that should the optimum precoding vectors is given described transmitter 710; Choose the decision rule of optimum precoding vectors, can follow formula (3).

In general, different precoding vectors can represent with different numerical value, only need feed back with the precoding vectors value corresponding of choosing just passablely during feedback, and do not need precoding vectors itself is fed back to transmitter.Certainly, the information of the described precoding vectors of present embodiment is not limited in top said corresponding numerical value, indicates selected precoding vectors as long as the content of feedback can be correct.

Described transmitter 710 comprises:

Code book list storage unit 711 is used for providing pre-coding matrix or precoding vectors to pre-coding matrix structural unit 712;

Pre-coding matrix structural unit 712 is used for the optimum precoding vectors information according to described receiver 720 decision units 723 feedbacks, chooses corresponding precoding vectors in code book list storage unit 711, the structure pre-coding matrix; Follow the principle of formula (4) according to precoding vectors structure pre-coding matrix.

Precoding unit 713 is used for the pre-coding matrix according to described pre-coding matrix structural unit 712 structures, carries out precoding.

Above embodiment three also provides the transmitter and receiver that is applied to the inventive method simultaneously.

For the described mthods, systems and devices of the specific embodiment of the invention, all within the spirit and principle of the specific embodiment of the invention, any modification of being done, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (10)

1. the method for precoding in the mimo system is characterized in that, adopts the quadrature Space Time Coding, and adopts two above data fluxions, and comprise:

In the beam shaping code book, choose optimum precoding vectors;

Adopt described optimum precoding vectors structure pre-coding matrix;

Adopt described pre-coding matrix to carry out precoding.

2. the method for precoding in the described mimo system of claim 1 is characterized in that, describedly chooses optimum precoding vectors in the beam shaping code book, follows following principle,

$v=\underset{v_i\∈\mathrm{\ζ}}{\arg \max }{\left|\right|H{v}_i\left|\right|}_{\text{2}}$

Wherein, v is optimum precoding vectors; v _iBe precoding vectors in the beam forming code book, i.e. the data fluxion code book that equals 1; H is a channel matrix; || || ₂Vector 2 norms are got in expression.

3. the method for precoding in the described mimo system of claim 2 is characterized in that, described beam forming code book is the beam forming code book in the IEEE802.16e agreement.

4. the method for precoding in the described mimo system of claim 1 is characterized in that, the described optimum precoding vectors structure pre-coding matrix of described employing is that n described optimum precoding vectors constituted pre-coding matrix side by side, and the value of n is identical with the data fluxion.

5. a mimo system that adopts limited feedback precoding comprises transmitter and receiver, it is characterized in that, described system adopts the quadrature Space Time Coding, and adopts two above data fluxions; And,

Described receiver comprises:

Channel estimating unit is used to carry out channel estimating, and channel estimation results is offered decision unit;

Code book list storage unit is used for providing pre-coding matrix or precoding vectors to decision unit;

Decision unit, the channel estimation results that provides according to described channel estimating unit and the decision rule of setting are provided, in the beam shaping code book of described code book list storage unit, choose optimum precoding vectors, and feedback information that should the optimum precoding vectors is given described transmitter;

Described transmitter comprises:

Code book list storage unit is used for providing pre-coding matrix or precoding vectors to the pre-coding matrix structural unit;

The pre-coding matrix structural unit is used for the optimum precoding vectors information according to described receiver decision unit feedback, chooses corresponding precoding vectors in the code book list storage unit of described transmitter, the structure pre-coding matrix;

Precoding unit is used for the pre-coding matrix according to described pre-coding matrix structural unit structure, carries out precoding.

6. the mimo system of the described employing limited feedback precoding of claim 5 is characterized in that, the information of the optimum precoding vectors of described feedback unit feedback is and each precoding vectors numerical value one to one.

7. a receiver that is used in the mimo system is characterized in that, described receiver comprises:

Channel estimating unit is used to carry out channel estimating, and channel estimation results is offered decision unit;

Code book list storage unit is used for providing pre-coding matrix or precoding vectors to decision unit;

Decision unit, the channel estimation results that provides according to described channel estimating unit and the decision rule of setting are provided, in the beam shaping code book of described code book list storage unit, choose optimum precoding vectors, and feedback information that should the optimum precoding vectors is given described transmitter;

8. the described receiver of claim 7 is characterized in that, described decision unit is chosen optimum precoding vectors and followed following principle in the beam shaping code book:

$v=\underset{v_i\∈\mathrm{\ζ}}{\arg \max }{\left|\right|H{v}_i\left|\right|}_{\text{2}}$

Wherein, v is optimum precoding vectors; v _iBe precoding vectors in the beam forming code book, i.e. the data fluxion code book that equals 1; H is a channel matrix; || || ₂Vector 2 norms are got in expression.

9. a transmitter that is used in the mimo system is characterized in that, described transmitter comprises:

Code book list storage unit is used for providing pre-coding matrix or precoding vectors to the pre-coding matrix structural unit;

The pre-coding matrix structural unit is used for the optimum precoding vectors information according to described receiver decision unit feedback, chooses corresponding precoding vectors in the code book list storage unit of described transmitter, the structure pre-coding matrix;

Precoding unit is used for the pre-coding matrix according to described pre-coding matrix structural unit structure, carries out precoding.

10. the described transmitter of claim 9 is characterized in that, described pre-coding matrix structural unit adopts described optimum precoding vectors information structuring pre-coding matrix, comprising:

According to described optimum precoding vectors information, determine optimum precoding vectors;

N described optimum precoding vectors constituted pre-coding matrix side by side, and the value of n is identical with the data fluxion.

Images