CN101227217B - Method and system for random wave packet forming based on multi-aerial receiver - Google Patents

Abstract

The invention discloses a randomized beam forming method and the system thereof on the basis of a multi-antenna receiver, which aims at limiting each receiver to obtain at most one beam, wherein each beam respectively carries out merger of maximum signal-to-noise ratio (SINR) by the receiver, and a feedback antenna which is merged obtains CSI of an equivalent channel. The system which is compared with the randomized beam forming system of a single antenna receiver is not added with the feedback cost of the system. A precoding mode, a CSI feedback content and a dispatching mode of the whole system are completely the same as the multi beam system of the single antenna receiver on the basis of a base station. The throughput of the randomized beam forming system is greatly increased under the premise that CSI feedback quantity is not added.

Description

Random beam forming method and system thereof based on multi-aerial receiver

Technical field

The present invention relates to wireless communication field, relate in particular to a kind of random beam forming system based on multi-aerial receiver.

Background technology

Multi-antenna technology (MIMO) can effectively utilize the radio communication multipath propagation environment, obtain the gain of spatial reuse (spatial multiplexing) and diversity (diversity), greatly improve capability of communication system to satisfy the capacity requirement of growing wireless communication system, be the technology of wireless communication system inevitable choice from now on, and be put in the wireless communication standard protocols such as 802.11n, 802.16e.

The analysis showed that of document 1, under multi-user multi-antenna (MU-MIMO) broadcast channel, channel capacity was with the MloglogKN speed increment when transmit leg was known complete channel information (full CSI).Wherein M is a number of transmit antennas, corresponding spatial reuse gain, and K is a number of users, and N is user's reception antenna number, and loglogKN is a multi-user diversity gain.And when transmit leg can't obtain channel information, channel capacity was only with the speed increment of logM.Therefore, the transmit leg CSI that obtains the user particularly plays important effect to the capacity of MU-MIMO system to mimo system.The method that reaches this ideal capacity is proposed by document 2, promptly utilizes stolen goods paper code (DPC) technology to realize at transmit leg.But the method operand is big, and is difficult to realize.Document 3 has proposed the ZF precoding technique, and document 4 has proposed vector disturbance precoding technique and come the approximation theory capacity.But the method in the above document all is to carry out under the base station obtains the prerequisite of complete CSI information.Base station in the practical wireless communication systems can only be realized by the CSI of receiver to base station feedback one definite form if seek out user's CSI.And multi-user wireless communication system is Limited Feedback (limited feedback) system, and the complete CSI of user feedback will bring huge expense to system.Therefore, the user reduce as far as possible feedback CSI with and improve throughput of system significance arranged.

The random beam forming technology is a kind of communication technology that collects feedback, is scheduling to one.Document 5 has proposed the thought of random beam forming the earliest.The random fading of channel has been introduced by many transmit antennas are carried out random weighting in the base station, thereby has increased the excursion and the rate of change of each user's equivalent channel fluctuation.When number of users is a lot, always there be certain user's channel and random wave bundle coupling, make its equivalent channel be in peak value.The base station scheduling user is transmitted the user when being in the channel top performance, has obtained multi-user's diversity gain, has improved power system capacity.But document 5 has only utilized wave beam transmission data, does not utilize the spatial reuse gain of multiaerial system.Document 6 has proposed the random beam forming method of multi-beam.Its operating process is that the base station produces a plurality of mutually orthogonal random wave bundles at random, and receiver is to the response of base station feedback to wave beam, and wave beam is distributed according to the SINR scheduling of user feedback again in the base station.Proved in the literary composition when number of users tends to infinite, can approach the base station with the method and obtain power system capacity under the complete CSI situation.The advantage of the method is: (1) user is less to the CSI of base station feedback amount, and each user only need feed back maximum SINR and the wave beam that reaches this SINR number, and need not to feed back concrete channel parameter, and the system feedback expense is little.(2) it is convenient that scheduling of resource is carried out in the base station.(3) precoding mode is simple, only wave beam and corresponding user data need be multiplied each other and can send.But the method just can be approached desirable channel capacity when number of users is tending towards infinite, and in actual cellular radio Communication system, number of users is limited, and throughput of system of realizing with the method for document 6 and desirable channel capacity differ greatly.

In order to improve the throughput of random beam forming system under less user situation, document 1 and document 6 have arrived the multi-receiver system with the single-receiver system extension.Receiver uses N root reception antenna, and has proposed 3 kinds of beam allocation and received signal processing mode.

Mode 1: regard every reception antenna as independent user, respectively to base station feedback CSI separately.This mode equivalence has increased user's number, has improved throughput of system, but feedback overhead is with reception antenna number linear growth.

Mode 2: limit each user and only distribute a wave beam, handle but received beam is not done further merging, then

${\mathrm{SINR}}_{k,m}=\frac{{\left(H_k{\mathrm{\φ}}_m\right)}^H\left(H_k{\mathrm{\φ}}_m\right)}{1/\mathrm{\ρ}+\underset{n=1,n\≠m}{\stackrel{M}{\mathrm{\Σ}}}{\left(H_k{\mathrm{\φ}}_n\right)}^H\left(H_k{\mathrm{\φ}}_n\right)}$

H wherein _kBe k user's matrix channel, φ _mM=1,2...M are random wave bundle, and ρ is a received signal to noise ratio.User k feedback

With corresponding wave beam number, $m_k^{*}=\underset{m=\mathrm{1,2}...M}{\arg \max }{\mathrm{SINR}}_{k,m}.$

Mode 3: with user is thread, or for user k distributes M/N wave beam, or do not distribute wave beam.The capacity C of each its support of user feedback _{K, m}

Document 7 has proposed the mode that receiver carries out day line options, and each receiver selects the CSI of an antenna of SINR maximum to feed back to the base station, and utilizes this antenna and reach the wave beam transmission data of this maximum SINR.The method is compared with the many antennas mode 1 in the document 6, because only feed back the CSI of an optimal antenna, has reduced feedback overhead, has guaranteed preferable performance simultaneously.

Chinese patent CN101090286 " wireless system, base station apparatus and terminal installation " proposes receiver and utilizes LMMSE filtering, eliminates the self-interference between each data flow.But its feedback quantity is linear increasing along with the increase of reception antenna.Chinese patent CN1855763 " the random wave bundle formation method that is used for multi-input multi-output system " proposes the base station and adjusts the wave beam number according to noise acoustic ratio situation self adaptation.When noise power is big, transmit with more wave beam.When noise power hour, reduce the number of wave beam.In a word, the thought of various receiving terminal processing methods is to improve the SINR of each wave beam as much as possible, to reach higher throughput of system.

Document 1:M.Sharif and B.Hassibi, " A comparison of time-sharing, DPC; andbeamforming for MIMO broadcast channels with many users, " IEEE Trans.Comm., vol.55, no.1, pp.11-15, Jan.2007.

Document 2:G.Caire and S.Shamai (Shitz), " On the achievable throughput of amulti-antenna Gaussian broadcast channel, " IEEE Trans.Info.Th., vol.49, no.7, pp.1691-1706, July 2003.

Document 3:Q.Spencer, A.L.Swindlehurst, and M.Haardt, " Zero-forcing methods fordownlink spatial multiplexing in multiuser MIMO channels; " IEEE Trans.Sig.Proc., vol.52, no.2, pp.462-471, Feb.2004.

Document 4:B.M.Hochwald, C.B.Peel, and A.L.Swindlehurst, " A vector-perturbationtechnique near capacity multiantenna multiuser communication-part II:perturbation; " IEEE Trans.Comm., vol.53, no.3, pp.537-544, March 2005.

Document 5:P.Viswanath, D.Tse, and R.Laroia, " Opportunistic beamforming usingdumb antennas, " IEEE Trans.Info.Th., vol.48, no.6, pp.1277-1294, June 2002.

Document 6:M.Sharif and B.Hassibi, " On the capacity of MIMO BC channel with partialside information, " IEEE Trans.Inf.Theory, vol.51, no.2, pp.506-522, Feb.2005.

Document 7:Lei Zan and Syed Ali Jafar, " Combined Opportunistic Beamforming andReceive Antenna Selection, " IEEE Communications Society WCNC, pp.1007-1011,2005

Summary of the invention

The purpose of this invention is to provide is not a kind ofly increasing the random beam forming method and the system thereof based on multi-aerial receiver that can improve throughput of system under the feedback overhead prerequisite.

For reaching above purpose, the main thought based on the random beam forming method of multi-aerial receiver that the present invention proposes is: utilize many antennas of receiver that each wave beam is done maximum Signal to Interference plus Noise Ratio (SINR) and merge, to improve the reception Signal to Interference plus Noise Ratio of all wave beams as much as possible, overcome inter-beam interference, and then improved throughput of system, limit the SINR after the merging that each receiver only feeds back an optimum beam simultaneously, guarantee that feedback quantity is identical with the random beam forming system of single reception antenna.Specifically may further comprise the steps:

1) base station produces wave beam information: produce M mutually orthogonal random wave bundle φ by the base station with M transmit antennas before each transmission of data blocks _m, m=1...M, M 〉=2;

2) receiver is done maximum Signal to Interference plus Noise Ratio merging to each wave beam: the receiver with N root reception antenna utilizes pilot tone to estimate N * Metzler matrix channel according to the LMMSE criterion, N 〉=2, and M 〉=2 are to each wave beam φ _m, m=1...M calculates the Signal to Interference plus Noise Ratio of each wave beam according to maximum Signal to Interference plus Noise Ratio criterion, and calculates the antenna combined vector of the wave beam that reaches maximum Signal to Interference plus Noise Ratio, and k user is to m wave beam φ _mMaximum Signal to Interference plus Noise Ratio SINR _{K, m}And combined vector x _{K, m}Calculate according to formula (1), (2):

${\mathrm{SINR}}_{k,m}={\left(H_k{\mathrm{\φ}}_m\right)}^H{B}_{k,m}^{-1}\left(H_k{\mathrm{\φ}}_m\right)---\left(1\right)$

$x_{k,m}=\frac{B_{k,m}^{-1}\left(H_k{\mathrm{\φ}}_m\right)}{{\left(H_k{\mathrm{\φ}}_m\right)}^H{B}_{k,m}^{-1}\left(H_k{\mathrm{\φ}}_m\right)}---\left(2\right)$

In the formula, H _kBe N * Metzler matrix channel of user k, φ _mBe m wave beam, $B_{k,m}=\mathrm{diag}(1/\mathrm{\ρ})+\underset{n\≠m}{\overset{M}{\underset{n=1}{\mathrm{\Σ}}}}{H}_k{\mathrm{\φ}}_n{\mathrm{\φ}}_n^H{H}_k^H,$ ρ is the signal to noise ratio of every reception antenna, and diag (1/ ρ) is the diagonal matrix of N * N of 1/ ρ for diagonal element, () ^HThe computing of expression conjugate transpose;

3) Signal to Interference plus Noise Ratio of the maximum of receiver after base station feedback merges and reach the wave beam number of this Signal to Interference plus Noise Ratio;

4) user's selection and beam allocation are carried out in the base station: the base station will select the user of same wave beam to be divided into one group, and note is done ${\mathrm{\Φ}}_m=\left\{k\right|m_k^{*}=m\},m=\mathrm{1,2}...M,$ The user who selects maximum Signal to Interference plus Noise Ratio from every group is with wave beam φ _mDistribute to this user k _m ^*

5) base station is with the message transmission rate C of user's support _Km*To user k _m ^*, m=1,2...M carries out data-modulated, and the wave beam that user data and user are got multiplies each other, and makes radio frequency processing after antenna emission, wherein C _Km*Calculate according to formula (3):

$C_{k_m^{*}}=\log (1+{\mathrm{SINR}}_{k_m^{*},m})---\left(3\right);$

6) user who is chosen by the base station merges and demodulation receiving data than combined vector according to its maximum letter is dry.

Realization the present invention is based on the system of the random beam forming method of multi-aerial receiver, comprises base station and a plurality of receiver, and the base station is by M root transmitting antenna, M 〉=2, random wave bundle generation module, pilot tone generation module, channel information demodulation module, scheduler module, modulation module, pretreatment module, radio-frequency module and base station transceiver duplex module are formed, it is characterized in that each receiver comprises N root reception antenna, N 〉=2; Receive the channel estimation module of the pilot signal of base station transmission, maximum letter is dry than combined vector computing module, to the maximum letter of base station feedback dry than and the feedback of channel information module of corresponding wave beam number, receiver control receives base station signal and receiver transmission duplex module to the base station feedback signal, to the received signal of each antenna according to dry post-processing module and the demodulation module that merges than combined vector computing module of maximum letter; The input of channel estimation module links to each other with the output of receiver transmission duplex module, an output of channel estimation module links to each other with the dry input than combined vector computing module of maximum letter, another output links to each other with an input of post-processing module, dry two outputs than combined vector computing module of maximum letter link to each other with input of feedback of channel information module and another input of post-processing module respectively, another input of feedback of channel information module links to each other with an output of demodulation module, and the output of feedback of channel information module links to each other with the input of receiver transmission duplex module; The output of post-processing module links to each other with the input of demodulation module, and another output of demodulation module is a user data.

Beneficial effect of the present invention is:

Utilize antenna merging method at receiver, improved the SINR of each wave beam, thereby improved the goodput of random beam forming system.System compares with single antenna receiver random beam forming, and native system does not increase the feedback overhead of system and the scheduling complexity of base station.

Description of drawings

Fig. 1 is the block diagram of system that adopts the inventive method;

Fig. 2 is the composition frame chart of the base station in the system of the present invention;

Fig. 3 is the composition frame chart of the receiver in the system of the present invention;

Maximum SINR antenna combined vector computing module calculating process schematic diagram when Fig. 4 is M=4;

Fig. 5 is transmitting antenna M=4, during reception antenna N=2, and under the multiple Gaussian channel condition of zero-mean unit variance, system of the present invention and document 6, the throughput performance curve chart of two kinds of antenna selection modes and single antenna receiver (N=1) system in the document 7.

Embodiment

The present invention is further elaborated below in conjunction with accompanying drawing.

As shown in Figure 1, the system of the random beam forming method based on multi-aerial receiver provided by the invention comprises base station and a plurality of receiver.

The base station as shown in Figure 2, by M root transmitting antenna 1, M 〉=2, random wave bundle generation module 2, pilot tone generation module 3, channel information (CSI) demodulation module 4, scheduler module 5, modulation module 6, data preprocessing module 7, radio-frequency module 8 and base station transceiver duplex module 9 are formed;

Wherein: random wave bundle generation module 2 before each transfer of data, produces the orthogonal random beam of M M dimension.Can produce the multiple gaussian random matrix of the independent same distribution zero-mean unit variance of M * M dimension earlier, decompose, get the Q matrix and obtain by this matrix being carried out QR.For reducing each calculating, also can store a large amount of orthogonal random beam groups in advance, therefrom select one group at every turn and be used for transmission;

Pilot tone generation module 3 is used to produce the sequence of pilot symbols of distinguishing M wave beam, simultaneously the sequence of also estimating as receiver channel;

CSI demodulation module 4, the maximum SINR of demodulate reception machine feedback and corresponding wave beam information;

Scheduler module 5 according to the CSI of each receiver feedback, with selecting the receiver of same beam number to be divided into one group, is selected the receiver of SINR maximum in every group, be used for transfer of data for this receiver beam allocation; The selection of receiver and beam allocation operation in scheduler module control modulating unit 6 and the data preprocessing module 7;

Modulating unit 6, the selected modulation system of supporting according to receiver of speed (as QPSK etc.) scheduler module 5 selected receiver data are modulated;

Data preprocessing module 7, under the control of scheduler module 5, modulating unit 6 is sent receiver data and the wave beam that distributes of scheduler module 5 multiply each other, send into radio-frequency module 8;

Radio-frequency module 8 to radiofrequency signal, is sent into base station transceiver duplex module 9 with the data up-conversion to be sent of data preprocessing module 7 output.

Base station transceiver duplex module 9, the control base station sends and received signal.

The native system receiver as shown in Figure 3, each receiver comprises N root reception antenna 10, N 〉=2, receive the channel estimation module 11 of the pilot signal of base station transmission, maximum letter is dry than combined vector computing module 12, to the maximum letter of base station feedback dry than and the feedback of channel information module 13 of corresponding wave beam number, receiver control receives base station signal and receiver transmission duplex module 14 to the base station feedback signal, received signal to each antenna is believed dry post-processing module 15 and the demodulation module 16 that merges than combined vector computing module 12 according to maximum, the input of channel estimation module 11 links to each other with the output of receiver transmission duplex module 14, an output of channel estimation module 11 links to each other with the dry input than combined vector computing module 12 of maximum letter, and another output links to each other with an input of post-processing module 15; Dry two outputs than combined vector computing module 12 of maximum letter link to each other with input of feedback of channel information module 13 and another input of post-processing module 15 respectively, another input of feedback of channel information module 13 links to each other with an output of demodulation module 16, the output of feedback of channel information module 13 links to each other with the input of receiver transmission duplex module 14, and the output of post-processing module 15 links to each other with the input of demodulation module 16; Another output of demodulation module 16 is a user data.

Below with base station number of transmit antennas M=4, it is example that each receiver reception antenna is counted N=2, and the inventive method is described, step is as follows:

1) base station produces wave beam information: utilize random wave bundle generation module 2 to produce 4 mutually orthogonal random wave bundle φ by the base station with 4 transmit antennas before each transmission of data blocks _m, m=1...4;

2) receiver is done maximum Signal to Interference plus Noise Ratio to each wave beam and merged: the channel estimation module 11 with receiver of 2 reception antennas utilizes pilot tone to estimate 2 * 4 matrix channel according to the LMMSE criterion, maximum letter dry than combined vector computing module 12 to each wave beam φ _m, m=1...4 calculates the Signal to Interference plus Noise Ratio of each wave beam according to maximum Signal to Interference plus Noise Ratio criterion, and calculates the antenna combined vector of the wave beam that reaches maximum Signal to Interference plus Noise Ratio, and k user is to m wave beam φ _mMaximum Signal to Interference plus Noise Ratio SINR _{K, m}And combined vector x _{K, m}Calculate according to formula (1), (2):

${\mathrm{SINR}}_{k,m}={\left(H_k{\mathrm{\φ}}_m\right)}^H{B}_{k,m}^{-1}\left(H_k{\mathrm{\φ}}_m\right)---\left(1\right)$

$x_{k,m}=\frac{B_{k,m}^{-1}\left(H_k{\mathrm{\φ}}_m\right)}{{\left(H_k{\mathrm{\φ}}_m\right)}^H{B}_{k,m}^{-1}\left(H_k{\mathrm{\φ}}_m\right)}---\left(2\right)$

In the formula, H _kBe 2 * 4 matrix channel of user k, φ _mBe m wave beam, $B_{k,m}=\mathrm{diag}(1/\mathrm{\ρ})+\underset{\stackrel{n=1}{n\≠m}}{\overset{M}{\mathrm{\Σ}}}{H}_k{\mathrm{\φ}}_n{\mathrm{\φ}}_n^H{H}_k^H,$ ρ is the signal to noise ratio of every reception antenna, and diag (1/ ρ) is 2 * 2 the diagonal matrix of 1/ ρ for diagonal element, () ^HThe computing of expression conjugate transpose;

Operating process below with reference to Fig. 4 Special attention will be given to receiver k antenna combined vector computing module 12.

A. to each wave beam φ _m, the H that obtains according to channel estimation module 11 _kφ m, m=1,2...4 calculates interfering beam plus noise autocorrelation matrix B according to formula (3) _{K, m}

$B_{k,m}=\underset{n\≠m}{\overset{M}{\underset{n=1}{\mathrm{\Σ}}}}\left(H_k{\mathrm{\φ}}_n\right){\left(H_k{\mathrm{\φ}}_n\right)}^H+\mathrm{diag}(1/\mathrm{\ρ})---\left(3\right)$

Wherein diag (1/ ρ) is 2 * 2 diagonal matrix of 1/ ρ for diagonal element;

B. calculate B _{K, m}Inverse matrix B _{K, m} ^-1

C. the SINR after according to formula (4) calculating receiver k wave beam m being merged _{K, m}

${\mathrm{SINR}}_{k,m}={\left(H_k{\mathrm{\φ}}_m\right)}^H{B}_{k,m}^{-1}\left(H_k{\mathrm{\φ}}_m\right)---\left(4\right)$

Among Fig. 4, the SINR of A1～A4 for obtaining after each wave beam is merged _{K, m}, m=1...4;

D. compare SINR _{K, m}, m=1,2...4 obtains optimum beam $m_k^{*}={\underset{m=1.2...M}{\arg \max }\mathrm{SINR}}_{k,m}$ With its corresponding SINR, ${\mathrm{SINR}}_{k,m_k^{*}}=\underset{m=1...M}{\max }{\mathrm{SINR}}_{k,m}.$ Simultaneously, according to formula (5) compute beam m _k ^*Maximum SINR antenna combined vector x _{K, mk*}

$x_{k,m_k^{*}}=\frac{B_{k,m_k^{*}}^{-1}\left(H_k{\mathrm{\φ}}_{m_k^{*}}\right)}{{\left(H_k{\mathrm{\φ}}_{m_k^{*}}\right)}^H{B}_{k,m_k^{*}}^{-1}\left(H_k{\mathrm{\φ}}_{m_k^{*}}\right)}---\left(5\right)$

Among Fig. 4, B ₁Be B _{K, mk*} ^-1(H _kφ _Mk*), B ₂Be SINR _{K, mk*}, C is x _{K, mk*}

3) receiver utilizes the Signal to Interference plus Noise Ratio of the maximum of feedback of channel information module 13 after base station feedback merges and reaches the wave beam number of this Signal to Interference plus Noise Ratio: receiver k calculates $m_k^{*}=\underset{m=1.2...M}{\arg \max }{\mathrm{SINR}}_{k,m}$ With ${\mathrm{SINR}}_{k,m_k^{*}}=\underset{m=1...M}{\max }{\mathrm{SINR}}_{k,m},$ To base station feedback (SINR _{K, mk*}, m _k ^*).

4) user's selection and beam allocation are carried out in the base station: the maximum SINR and the corresponding wave beam information of the CSI demodulation module 4 demodulate reception machines feedback of base station, scheduler module 5 will select the user of same wave beam to be divided into one group, the note work ${\mathrm{\Φ}}_m=\left\{k\right|m_k^{*}=m\},m=\mathrm{1,2}...4,$ From every group, select the user of maximum Signal to Interference plus Noise Ratio $k_m^{*}=\underset{k\∈{\mathrm{\Φ}}_m}{\arg \max }{\mathrm{SINR}}_{k,m_k^{*}},$ With wave beam φ _mDistribute to this user k _m ^*

5) modulating unit 6 of base station is with the message transmission rate C of user's support _Km*To user k _m ^*, m=1,2...4 carries out data-modulated, and data preprocessing module 7 multiplies each other the wave beam that user's modulating data and user gets, and utilizes radio-frequency module 8 to make radio frequency processing after antenna emission, wherein C _Km*Calculate according to formula (6):

$C_{k_m^{*}}=\log (1+{\mathrm{SINR}}_{k_m^{*},m})---\left(6\right);$

6) user who is chosen by the base station utilizes post-processing module 15 to merge receiving data than combined vector according to its maximum letter is dry, does data demodulates through demodulation module 16 and recovers original user data.

Fig. 5 is transmitting antenna M=4, during reception antenna N=2, under the multiple Gaussian channel environment of zero-mean unit variance, system of the inventive method and document 6 (N=2, feed back all antenna CSI), the throughput of system performance chart in the document 7 (N=2, the CSI of the antenna that feedback is optimum) when two kinds of antenna selection modes and single antenna receiver N=1.Dotted line is M=4 among the figure, the theoretical progressive capacity curve of the random beam forming system of N=2, as seen the throughput that the throughput that antenna merging method of the present invention obtains antenna selecting method under the identical reception antenna number obtains, power system capacity also can reach the effect more than the 5bit/s/Hz when less receiver, and does not increase feedback quantity.In the base station, the precoding mode of whole system, CSI feedback content, scheduling mode and single antenna receiver random beam forming system are identical, under the prerequisite that does not increase the CSI feedback quantity, have improved the throughput of random beam forming system greatly.

Claims (2)

1. based on the random beam forming method of multi-aerial receiver, may further comprise the steps:

1) base station produces wave beam information: produce M mutually orthogonal random wave bundle φ by the base station with M transmit antennas before each transmission of data blocks _m, m=1...M, M 〉=2;

2) receiver is done maximum Signal to Interference plus Noise Ratio merging to each wave beam: the receiver with N root reception antenna utilizes pilot tone to estimate N * Metzler matrix channel according to the LMMSE criterion, N 〉=2, and M 〉=2 are to each wave beam φ _m, m=1...M calculates the Signal to Interference plus Noise Ratio of each wave beam according to maximum Signal to Interference plus Noise Ratio criterion, and calculates the antenna combined vector of the wave beam that reaches maximum Signal to Interference plus Noise Ratio, and k user is to m wave beam φ _mMaximum Signal to Interference plus Noise Ratio SINR _{K, m}And combined vector x _{K, m}Calculate according to formula (1), (2):

${\mathrm{SINR}}_{k,m}={\left(H_k{\mathrm{\φ}}_m\right)}^H{B}_{k,m}^{-1}\left(H_k{\mathrm{\φ}}_m\right)---\left(1\right)$

$x_{k,m}=\frac{B_{k,m}^{-1}\left(H_k{\mathrm{\φ}}_m\right)}{{\left(H_k{\mathrm{\φ}}_m\right)}^H{B}_{k,m}^{-1}\left(H_k{\mathrm{\φ}}_m\right)}---\left(2\right)$

In the formula, H _kBe N * Metzler matrix channel of user k, φ _mBe m wave beam,

ρ is the signal to noise ratio of every reception antenna, and diag (1/ ρ) is the diagonal matrix of N * N of 1/ ρ for diagonal element, () ^HThe computing of expression conjugate transpose;

3) Signal to Interference plus Noise Ratio of the maximum of receiver after base station feedback merges and reach the wave beam number of this Signal to Interference plus Noise Ratio;

4) user's selection and beam allocation are carried out in the base station: the base station will select the user of same wave beam to be divided into one group, and note is done

From every group, select the user of maximum Signal to Interference plus Noise Ratio

With wave beam φ _mDistribute to this user

5) base station is with the message transmission rate of user's support

To the user

Carry out data-modulated, the wave beam that user data and user are got multiplies each other, and makes radio frequency processing after the antenna emission, wherein

Calculate according to formula (3):

$C_{k_m^{*}}=\log (1+\mathrm{SIN}{R}_{k_m^{*},m})---\left(3\right);$

6) user who is chosen by the base station merges and demodulation receiving data according to its maximum Signal to Interference plus Noise Ratio combined vector.

2. the system of the random beam forming method based on multi-aerial receiver according to claim 1, comprise base station and a plurality of receiver, the base station is by user data, M root transmitting antenna (1), M 〉=2, random wave bundle generation module (2), pilot tone generation module (3), channel information demodulation module (4), scheduler module (5), modulation module (6), pretreatment module (7), radio-frequency module (8) and base station transceiver duplex module (9) are formed, the pilot tone preface that pilot transmission module (3) produces is carried out beam forming by random wave bundle generation module (2) and is defeated by radio-frequency module (8), the output of radio-frequency module (8) links to each other with the input of transmission duplex module (9), the output of transmission duplex module (9) links to each other with transmitting antenna (1), transmission duplex module (9) will be carried out demodulation from the feedback signal input CSI demodulation module (4) of receiver, the restituted signal input scheduling module (5) of CSI demodulation module (4) output, the control signal of scheduler module (5) output is imported user data respectively, modulation module (6) and pretreatment module (7), the output of user data links to each other with modulation module (6), the modulation signal input pretreatment module (7) that modulation module (6) produces is carried out preliminary treatment, the output of pretreatment module (7) links to each other with the input of radio-frequency module (8), it is characterized in that each receiver comprises N root reception antenna (10), N 〉=2; Receive the channel estimation module (11) of the pilot signal of base station transmission, maximum Signal to Interference plus Noise Ratio combined vector computing module (12), feedback of channel information module (13) to the maximum Signal to Interference plus Noise Ratio of base station feedback and corresponding wave beam number, receiver control receives base station signal and receiver transmission duplex module (14) to the base station feedback signal, post-processing module (15) and demodulation module (16) that the received signal of each antenna is merged according to maximum Signal to Interference plus Noise Ratio combined vector computing module (12), the input of channel estimation module (11) links to each other with the output of receiver transmission duplex module (14), an output of channel estimation module (11) links to each other with the input of maximum Signal to Interference plus Noise Ratio combined vector computing module (12), another output links to each other with an input of post-processing module (15), two outputs of maximum Signal to Interference plus Noise Ratio combined vector computing module (12) link to each other with input of feedback of channel information module (13) and another input of post-processing module (15) respectively, another input of feedback of channel information module (13) links to each other with an output of demodulation module (16), and the output of feedback of channel information module (13) links to each other with the input of receiver transmission duplex module (14); The output of post-processing module (15) links to each other with the input of demodulation module (16), and another output of demodulation module (16) is a user data.

Images