CN101136718A - Multi-input multi-output space multiplexing precoding method of wireless communication system - Google Patents

Abstract

Using precoded space multiplexing of limited fed back, the method defines a group of set of precoded matrix (i.e. precoded codebook) known by transmitter and receiver. Using algorithm of channel estimation to obtain current state information of channel, the receiver selects an optimal precoded matrix from precoded codebooks according to minimized principles of trace of mean square error matrix as well as through a feedback channel, feeds index of the precoded matrix back to the transmitter within coherence time of channel. Based on the said index of the precoded matrix, the transmitter obtains the precoded matrix, and then carries out precoded operations on vectors of transmit signal based on the precoded matrix. Precoded codebook can be designed based on maximized principle of minimum distance between two norms of projections of subspaces. Using limited bit number fed back, the invention raises communication performance.

Description

The method for precoding of the spatial reuse of multiple-input and multiple-output in the wireless communication system

Technical field

The present invention relates to the multi-antenna technology field in the radio communication, relate in particular to the space multiplexing technique of precoding in the wireless communication system that uses multiple-input and multiple-output MIMO (Multiple Input Multiple output).

Background technology

The MIMO technology outstanding feature is exactly greatly to improve the spectrum efficiency (or power system capacity) of system and the reliability of Radio Link in the wireless communication field.Can derive based on information theory, if the transmitting power all identical (supposing that promptly transmitting terminal does not have priori to channel matrix) that each transmitting antenna is total, the MIMO capability of communication system is with min{Mt, the Mr} linear growth, and wherein Mt, Nr are respectively and transmit and receive antenna number.Simultaneously, some scholars propose, if the complete known or part known channel matrix of transmitting terminal, and make full use of the power that known channel matrix information is adjusted each transmitting antenna of MIMO communication system, then can further improve the MIMO capability of communication system.Therefore, in order to make the MIMO capability of communication system big as far as possible, a lot of patents have proposed the mimo system of closed loop precoding, be that all or part of feedback of channel information that the receiver of MIMO communication system will be estimated is given transmitter, the channel information that the transmitter utilization receives carries out precoding to transmitting.

In the wireless communication system of multiple-input and multiple-output (MIMO), spatial reuse (spatialmultiplexing) is divided into the independently parallel subsequence in several roads and transmits simultaneously by importing data, thereby obtained higher frequency utilization efficient, yet, and be subject to the influence of mimo channel rank of matrix deficiency because multiplex techniques does not have redundancy on the spatial domain.Precoding technique at first will send sequence and be divided into M parallel subsequence, be assigned to different transmitting antennas again with after pre-coding matrix multiplies each other, thereby can suppress the influence of mimo channel rank of matrix deficiency to systematic function effectively.

Research to precoding, mostly based on complete this assumed condition of known channel state information of transmitting terminal, yet in a lot of mimo systems, the uplink and downlink channel does not have reciprocity, in the FDD system, this moment, transmitting terminal can only obtain the state information of channel by feedback channel, yet because limited frequency spectrum resources only can provide limited feedback bit number at the internal feedback coherence time of channel channel.How utilizing limited feedback bit number to improve communication performance in this case, is a practicality and challenging research topic.

Summary of the invention

Technical problem to be solved by this invention is to provide the method for precoding of the spatial reuse of multiple-input and multiple-output in a kind of wireless communication system, improves communication performance to utilize limited feedback bit number.

For solving the problems of the technologies described above, the invention provides the method for precoding of the spatial reuse of multiple-input and multiple-output in a kind of wireless communication system, comprise the steps:

(1) set of the pre-coding matrix that pre-defined one group of transmitter and receiver is all known;

(2) receiver is selected a pre-coding matrix according to the current channel information of estimating to obtain from the set of described predefined pre-coding matrix;

(3) feedback information of the pre-coding matrix that will select of receiver is to transmitter;

(4) transmitter is determined the pre-coding matrix of current selection according to the information of the pre-coding matrix of the described selection that receives;

(5) transmitter carries out pre-encode operation according to the pre-coding matrix of described current selection to emission signal vector.

Wherein, the set of the described pre-coding matrix of step (1), be with the maximized principle of two norm minimum ranges of the projection of subspace and by pre-defined, or with the maximized principle predefine of the Fubini-Study minimum range of subspace, or with the maximized principle predefine of the string minimum range of subspace.Wherein, under the situation of 4 feedback bits of 2 reception antennas of 4 transmit antennas, the set of transmitter and receiver predefine pre-coding matrix is a code book at the tenth of the twelve Earthly Branches, and the column vector of any one pre-coding matrix wherein is mutually orthogonal.

Wherein, in the described step (2), described receiver estimates to obtain current channel information according to pilot tone.

Wherein, in the described step (2), described current channel information comprises the channel condition information of current channel.

Wherein, in the described step (2), described current channel information comprises the signal to noise ratio information of current channel.

Wherein, in the described step (2), the described step of from the set of predefined pre-coding matrix, selecting a pre-coding matrix, be to minimize or maximum capacity or the minimized principle selection of mean square error determinant of a matrix, from described pre-coding matrix set, select the pre-coding matrix of an optimum according to the mean square error matrix trace.

Wherein, in the described step (3), described receiver arrives transmitter by the feedback information of the pre-coding matrix that feedback channel will be selected.

Wherein, in the described step (3), described receiver arrives transmitter in the feedback information of the pre-coding matrix that will select the coherence time of channel.

Wherein, in described step (3), the step (4), the information of the pre-coding matrix of described selection is the index of this pre-coding matrix.

The method for precoding of the spatial reuse of the multiple-input and multiple-output that the present invention realizes in wireless communication system, only utilizes limited feedback bit number to improve communication performance.

Description of drawings

Fig. 1 is according to the described realization flow figure of embodiments of the invention;

Fig. 2 is the simulation result of the link performance of the spatial reuse of use precoding in the non-coded system.

Embodiment

This invention is intended to provide a kind of method of in wireless telecommunication system, using the precoding of closed-loop MIMO, the method of the spatial reuse of the precoding of employing Limited Feedback, the set (being precoding codebook) of the pre-coding matrix that promptly pre-defined one group of transmitter and receiver is all known; Receiver obtains current state information of channel by channel estimation method, from the code book of precoding, select the pre-coding matrix of an optimum according to the minimized principle of mean square error matrix trace, and in the coherence time of channel, the index of this pre-coding matrix is fed back to transmitter by feedback channel, transmitter obtains pre-coding matrix according to the index of this pre-coding matrix, according to this pre-coding matrix emission signal vector is carried out pre-encode operation then; Precoding codebook is the code book that designs precoding with the maximized principle of two norm minimum ranges of the projection of subspace, perhaps with the predefined code book of the maximized principle of Fubini-Study (naming) minimum range of subspace, perhaps with the predefined code book of the maximized principle of string minimum range of subspace with name.

The present invention has simultaneously provided under this optimum code book design principle, concrete optimum precoding codebook under the situation of 4 feedback bits of 2 reception antennas of 4 transmit antennas, the simulation result proof has certain performance gain by the code book of this principle design than present code book.

Particular content of the present invention can be described below:

Thought of the present invention is that the emission pre-coding matrix is selected from a limited pre-coding matrix set.This group pre-coding matrix is that transmitter and receiver is known, and we are called precoding codebook.Receiver obtains current state information of channel by channel estimation method, select the pre-coding matrix of an optimum according to certain selection criterion, and the index of this pre-coding matrix is fed back to transmitter by feedback channel, transmitter obtains pre-coding matrix (precoding codebook is stored in the transmitter and receiver in advance) according to the index of this pre-coding matrix, according to this pre-coding matrix emission signal vector is carried out pre-encode operation then.

At transmitting terminal, bit stream is become M road stream of modulation symbols by demultiplexing (perhaps being called serial/parallel conversion) then through after encoding, modulating; An alternative scheme is that bit stream is become M road bit stream by demultiplexing (perhaps being called serial/parallel conversion) earlier, and each road becomes M road stream of modulation symbols after encoding, modulating then.At moment k, produced symbolic vector S like this _k=[S _{K, 1}S _{K, 2}... S _{K, M}] ^TThis symbolic vector S _kMultiply by a M then _tIt is the vector of Mt that XM pre-coding matrix F (selection of this pre-coding matrix can be described later) produces length $X_k=\sqrt{\frac{E_s}{M}}{\mathrm{FS}}_k,$ Here E _sIt is total transmitting power.If used the OFDM modulation, so every road symbols streams is carried out the operation of corresponding OFDM again, launches by many antennas again; If do not use OFDM, then directly launch by many antennas.

At receiving terminal, the baseband equivalence signal of reception is:

$Y_k=\sqrt{\frac{E_s}{M}}{\mathrm{HFS}}_k+V_k$ , H is a channel matrix here, V _kIt is noise vector.

Use MMSE (minimum mean-square error) receiver to decipher at receiving terminal.At this, be example with linear MMSE receiver: $G={[F^H{H}^H\mathrm{HF}+\left(\frac{{\mathrm{MN}}_0}{E_s}\right)I_M]}^{-1}{F}^H{H}^H.$

Decipher by following form at receiver end:

${\hat{S}}_k=G{Y}_k$

The channel matrix H here and

Can obtain by channel estimation method and signal-to-noise ratio estimation algorithm at receiving terminal by pilot tone.

Though receiver is that example describes with linear MMSE, is not limited to linear MMSE, and is suitable equally for the MMSE receiver of other types, for example iteration MMSE receiver.

At receiving terminal, the present invention selects pre-coding matrix (being code word) according to following principle in precoding codebook:

$F=\underset{F_I\∈F}{\arg \min \mathrm{trace}}\left(\mathrm{MSE}\left(F_i\right)\right)$

$\mathrm{trace}\left(\mathrm{MSE}\left(F\right)\right)=\mathrm{trace}\left(\frac{E_s}{M}{[\left(\frac{E_s}{{\mathrm{MN}}_0}\right)F^H{H}^H\mathrm{HF}+I_M]}^{-1}\right):$

Wherein:

The mark of trace representing matrix (being the diagonal element sum).

For the wireless telecommunication system that uses OFDM,

$\mathrm{trace}\left(\mathrm{MSE}\left(F\right)\right)=\underset{j}{\mathrm{\Σ}}\mathrm{trace}\left(\frac{E_s}{M}{[\left(\frac{E_s}{{\mathrm{MN}}_0}\right)F^H{H_j}^H{H}_{j}F+I_M]}^{-1}\right),$ H wherein _jThe channel coefficient matrix of representing j subcarrier.

The precoding codebook all known in advance to transmitter, receiver designs according to following principle: select to make the code book of minimum value maximum in projection two norms of column space of any a pair of code word matrix (pre-coding matrix).The design that is precoding codebook F should maximize

Here d _P2(F _i, F _j) expression F _iAnd F _jProjection two norms of the subspace of being opened are defined as follows:

$d_{p2}(F_i,F_j)={||F_i{F}_j^H-F_j{F}_i^H||}_2$

According to the design of this principle, as follows for the code book of the space multiplexing system optimum of the closed loop of 2 reception antenna 4 bit feedback of 4 transmit antennas:

The code book of 2 reception antenna 4 bit feedback of table 1:4 transmit antennas

Pre-coding matrix index number	Column1	Column2
			1	0.0962+0.4648i	0.2347-0.6270i
0.2782+0.6514i	-0.0843-0.0245i
		-0.4252-0.2135i		-0.2324-0.4021i
-0.2156	-0.5731
		2		0.0191-0.0914i	0.2654+0.3875i
-0.1905+0.2672i	0.5781-0.5601i
			-0.0130+0.2955i	-0.0510-0.0784i
0.8922	0.3504
			3	0.2122-0.2470i	0.1605-0.6748i
-0.2191+0.2717i	0.2446-0.1753i
		0.3803+0.0452i		0.3978+0.3933i
0.7908	-0.3397
		4		-0.1473+0.5963i	-0.0588+0.0246i
-0.3022+0.2338i	0.2828-0.4373i
			0.0940+0.3170i	0.1248-0.6161i
0.6061	0.5741
			5	-0.1746-0.6538i	0.3520+0.4416i
0.4352-0.1163i	-0.0459-0.3497i
		-0.0629+0.1223i		0.1256-0.3156i
0.5659	0.6643
		6		0.1884-0.8249i	-0.1363+0.0803i
0.0684-0.4147i	0.2392+0.4012i
			-0.0119-0.0658i	-0.4829-0.2019i
0.3208	0.6949
			7	-0.0638+0.7263i	-0.4474-0.3599i
-0.0019-0.3922i	-0.1876-0.0408i
		-0.4584+0.0946i		0.0055+0.5974i
0.3091	0.5259
		8		-0.3021+0.0419i	0.6658-0.1688i

	0.0253-0.0299i	-0.3959-0.5629i
			-0.5238+0.6615i	-0.0908+0.1090i
	-0.4399	-0.1858
			9	0.5049-0.0156i	0.1661+0.4162i
-0.0721+0.1338i	0.0001-0.0592i
		-0.4751-0.2763i		0.3718+0.6728i
0.6478	0.4525
		10		0.4704+0.1170i	0.1461+0.1383i
0.4108-0.0525i	-0.7159+0.5112i
			-0.6526-0.3397i	-0.4100+0.1242i
0.2288	0.0467
			11	-0.6454+0.2182i	0.0394+0.2796i
0.4437+0.1299i	-0.6793-0.0264i
		-0.2522-0.0133i		-0.3539-0.4647i
0.5083	0.3418
		12		0.0806+0.1836i	0.0954+0.0929i
-0.1701-0.2366i	0.4952+0.4922i
			0.0921+0.8482i	0.0159-0.1082i
-0.3832	-0.6949
			13	-0.4611-0.3301i	0.0644+0.0628i
0.3125-0.0822i	-0.2357+0.6715i
		-0.0444+0.4814i		0.4004-0.2255i
0.5834	0.5237
		14		-0.1412-0.4628i	-0.0714-0.3358i
-0.1229+0.2240i	0.0195-0.4623i
			-0.1449+0.5543i	0.2737-0.5859i
0.6102	0.4998
			15	0.8628+0.0608i	0.1373-0.2705i
0.1006+0.3005i	-0.8524+0.0032i
		0.3175-0.0168i		-0.2812-0.0314i
0.2247	0.3185
		16		-0.4669-0.0392i	0.2133+0.2619i

	-0.1828+0.1433i	-0.820l+0.0924i
			-0.2339-0.6949i	0.1072+0.2616i
	-0.4346	-0.3533

Precoding codebook shown in the table 1 has following characteristics: this precoding codebook is a code book at the tenth of the twelve Earthly Branches, and the column vector of any one pre-coding matrix wherein is mutually orthogonal, and wherein column vector is the mutually orthogonal following condition that satisfies:

F _g ^HF _g=I _M, g=1 ..., G, G=2 ^BB is the bit number of feedback.

Receiver according to $F_{\mathrm{select}}=\underset{F_i\∈F}{\arg \min \mathrm{trace}}\left(\mathrm{MSN}\left(F_i\right)\right)$ The principle of (the mean square error matrix trace minimizes) (here, also can be maximum capacity principle or mean square error determinant of a matrix minimization principle) pre-coding matrix of selection from predefined code book, and in the coherence time of channel, the index of this pre-coding matrix is fed back to transmitter by feedback channel, transmitter obtains pre-coding matrix (precoding codebook is stored in the transmitter and receiver in advance) according to the index of this pre-coding matrix, according to this pre-coding matrix emission signal vector is carried out pre-encode operation then, thereby form the process of spatial reuse of the precoding of a closed loop.

As shown in Figure 1, for according to the described realization flow figure of embodiments of the invention.Comprise the steps:

The maximized principle of two norm minimum ranges of pressing the projection of subspace defines the code book (step 101) of the precoding of an optimization;

Transmitter carries out Space Time Coding (step 102) to symbols streams;

Transmitter is selected corresponding pre-coding matrix according to pre-coding matrix index, and carries out corresponding pre-encode operation (step 103);

Receiver is selected the pre-coding matrix (step 104) an of the best from the code book of precoding according to the minimized principle of mean square error matrix trace;

Receiver feeds back to transmitter (step 105) with the index of this pre-coding matrix by feedback channel in the coherence time of channel.

As shown in Figure 2, be the simulation result of the link performance of the spatial reuse of use precoding in the non-coded system, its simulated conditions is as follows:

Bandwidth: 5MHz;

Frequency: 2GHz;

The OFDM relevant parameter:

Subcarrier spacing: 9.6KIdz;

FFT counts: 512;

Useful subcarrier number: 480;

Protection subcarrier number: 32;

Antenna configurations: 4 transmit antennas, 2 reception antennas;

Channel model: the SCM channel model of simplification;

Receiver: linear MMSE receiver;

Channel estimating: ideal communication channel is estimated.

Document shown in Fig. 2 [1] is: 3GPP2 TSG-C WG3 ContributionC30-20060731-039R2, " Fourier-based MIMO Precoders ";

Shown document [2] is: 3GPP2 TSG-C WG3 Contribution C30-20060911-072, " MIM0 Design for LBC-FDD ".

Claims (12)

1. the method for precoding of the spatial reuse of multiple-input and multiple-output in the wireless communication system is characterized in that, comprises the steps:

(1) set of the pre-coding matrix that pre-defined one group of transmitter and receiver is all known;

(2) receiver is selected a pre-coding matrix according to the current channel information of estimating to obtain from the set of described predefined pre-coding matrix;

(3) feedback information of the pre-coding matrix that will select of receiver is to transmitter;

(4) transmitter is determined the pre-coding matrix of current selection according to the information of the pre-coding matrix of the described selection that receives;

(5) transmitter carries out pre-encode operation according to the pre-coding matrix of described current selection to the emission signal vector through Space Time Coding.

2. the method for claim 1, it is characterized in that, the set of the described pre-coding matrix of step (1), be with the maximized principle of two norm minimum ranges of the projection of subspace and by pre-defined, or with the maximized principle predefine of the Fubini-Study minimum range of subspace, or with the maximized principle predefine of the string minimum range of subspace.

3. the method for claim 1 is characterized in that, described step (1) comprising: the set of described pre-coding matrix is stored in the transmitter and receiver in advance.

4. the method for claim 1 is characterized in that, in the described step (2), described receiver estimates to obtain current channel information according to pilot tone.

5. the method for claim 1 is characterized in that, in the described step (2), described current channel information comprises the channel condition information of current channel.

6. the method for claim 1 is characterized in that, in the described step (2), described current channel information comprises the signal to noise ratio information of current channel.

7. the method for claim 1, it is characterized in that, in the described step (2), the described step of selecting a pre-coding matrix from the set of predefined pre-coding matrix minimizes or maximum capacity or the minimized principle selection of mean square error determinant of a matrix according to the mean square error matrix trace.

8. the method for claim 1 is characterized in that, in the described step (3), described receiver arrives transmitter by the feedback information of the pre-coding matrix that feedback channel will be selected.

9. the method for claim 1 is characterized in that, in the described step (3), described receiver arrives transmitter in the feedback information of the pre-coding matrix that will select the coherence time of channel.

10. the method for claim 1 is characterized in that, in described step (3), the step (4), the information of the pre-coding matrix of described selection is the index of this pre-coding matrix.

11. method as claimed in claim 2, it is characterized in that, under the situation of 4 feedback bits of 2 reception antennas of 4 transmit antennas, transmitter and receiver predefine precoding codebook is a code book at the tenth of the twelve Earthly Branches, and the column vector of any one pre-coding matrix wherein is mutually orthogonal.

12. method as claimed in claim 11 is characterized in that, under the situation of 4 feedback bits of 2 reception antennas of 4 transmit antennas, transmitter and receiver predefine pre-coding matrix is:

Pre-coding matrix index number Column1 Column2 1 0.0962+0.4648i 0.2347-0.6270i 0.2782+0.6514i -0.0843-0.0245i -0.4252-0.2135i -0.2324-0.4021i -0.2156 -0.5731 2 0.0191-0.0914i 0.2654+0.3875i -0.1905+0.2672i 0.5781-0.5601i -0.0130+0.2955i -0.0510-0.0784i 0.8922 0.3504 3 0.2122-0.2470i 0.1605-0.6748i -0.2191+0.2717i 0.2446-0.1753i 0.3803+0.0452i 0.3978+0.3933i 0.7908 -0.3397 4 -0.1473+0.5963i -0.0588+0.0246i -0.3022+0.2338i 0.2828-0.4373i 0.0940+0.3170i 0.1248-0.6161i

0.6061 0.5741 5 -0.1746-0.6538i 0.3520+0.4416i 0.4352-0.1163i -0.0459-0.3497i -0.0629+0.1223i 0.1256-0.3156i 0.5659 0.6643 6 0.1884-0.8249i -0.1363+0.0803i 0.0684-0.4147i 0.2392+0.4012i -0.0119-0.0658i -0.4829-0.2019i 0.3208 0.6949 7 -0.0638+0.7263i -0.4474-0.3599i -0.0019-0.3922i -0.1876-0.0408i -0.4584+0.0946i 0.0055+0.5974i 0.3091 0.5259 8 -0.3021+0.0419i 0.6658-0.1688i 0.0253-0.0299i -0.3959-0.5629i -0.5238+0.6615i -0.0908+0.1090i -0.4399 -0.1858 9 0.5049-0.0156i 0.1661+0.4162i -0.0721+0.1338i 0.0001-0.0592i -0.4751-0.2763i 0.3718+0.6728i 0.6478 0.4525 10 0.4704+0.1170i 0.1461+0.1383i 0.4108-0.0525i -0.7159+0.5112i -0.6526-0.3397i -0.4100+0.1242i 0.2288 0.0467 11 -0.6454+0.2182i 0.0394+0.2796i 0.4437+0.1299i -0.6793-0.0264i -0.2522-0.0133i -0.3539-0.4647i 0.5083 0.3418 12 0.0806+0.1836i 0.0954+0.0929i -0.1701-0.2366i 0.4952+0.4922i 0.0921+0.8482i 0.0159-0.1082i

-0.3832 -0.6949 13 -0.4611-0.3301i 0.0644+0.0628i 0.3125-0.0822i -0.2357+0.6715i -0.0444+0.4814i 0.4004-0.2255i 0.5834 0.5237 14 -0.1412-0.4628i -0.0714-0.3358i -0.1229+0.2240i 0.0195-0.4623i -0.1449+0.5543i 0.2737-0.5859i 0.6102 0.4998 15 0.8628+0.0608i 0.1373-0.2705i 0.1006+0.3005i -0.8524+0.0032i 0.3175-0.0168i -0.2812-0.0314i 0.2247 0.3185 16 -0.4669-0.0392i 0.2133+0.2619i -0.1828+0.1433i -0.8201+0.0924i -0.2339-0.6949i 0.1072+0.2616i -0.4346 -0.3533

Images