CN102111245B - Adaptive limiting feedback method and user equipment - Google Patents

Abstract

The invention discloses an adaptive limiting feedback method and user equipment (UE). The UE according to the invention comprises a receiving unit, a channel estimation unit, a feedback bit number selecting unit, a code word matching unit and a feedback unit, wherein the receiving unit is used for receiving average channel information broadcasted by a base station; the channel estimation unit is used for implementing channel estimation to obtain current channel status information; the feedback bit number selecting unit selects the feedback bit number Bk used for feeding back channel direction information according to the average channel information broadcasted by the base station and the obtained current channel status information; the code word matching unit is used for searching the most matching code word in first 2<Bk> code words of a code book according to a predetermined code word matching criterion; and the feedback unit feeds back the sequence number of the searched most matching code word, which is taken as the channel direction information, to the base station. According to the invention, the feedback overhead of the system channel direction information is reduced, and meanwhile, the system throughput is effectively increased. In addition, the user equipment can be well compatible with the current orthogonal beam forming technology, thereby having very high application value.

Description

Self adaptation limited feedback method and subscriber equipment

Technical field

The present invention relates to multiple-input and multiple-output (MIMO) technical field of radio communication, specifically channel state information feedback method and the subscriber equipment in a kind of multiuser MIMO quadrature space division multiplexing (OSDM) system.

Background technology

Limited Feedback is to solve the key technology of client feeds back channel condition information in the descending multi-user mimo system.How effectively to reduce the user feedback expense, guarantee that simultaneously power system capacity is a main study hotspot in current space division multi-access.

In traditional descending multi-user mimo system, each user's number of bits of feedback is fixed.It is generally acknowledged that the raising number of bits of feedback can make the channel condition information of feedback more accurate, thereby improve user rate, but when in system, certain user's channel is relatively relatively poor, improve number of bits of feedback also not obvious for this user's speed lifting.Therefore adopt fixing number of bits of feedback must cause overhead waste to a certain degree.

In the article " Multi-Antenna Downlink Channels withLimited Feedback and User Selection " (referring to " Selected Areas inCommunications; IEEE Journal on Volume 25; Issue 7; September2007Page (s): 1478-1491. ") of Taesang Yoo, when having provided employing orthogonal beams forming technique, user SINR and channel errors contacting between both.

at the article " Antenna Combining for the MIMO DownlinkChannel " of Niha Jindal (referring to " Wireless Communications, IEEE Transactions onVolume 7, Issue 10, October 2008 Page (s): 3834-3844. ") in, studied when system's code book produces at random (RVQ), relation between number of bits of feedback and channel directional information Channel Direction Information (CDI) quantization error, and a kind of scaling feedback scheme proposed, this scheme can cause larger feedback overhead when base station and user side antenna number differ larger, and the user can't adjust number of bits of feedback according to residing system performance self adaptation.

Summary of the invention

The present invention is based on above-mentioned two pieces of articles of Taesang Yoo and Niha Jindal, a kind of user's self adaptation Limited Feedback thought based on mimo system has been proposed, derive the optimal policy of user's self adaptation Limited Feedback, and a kind of self adaptation Limited Feedback scheme that more is close to real system proposed, its characteristics are: the initial access phase of user feeds back with identical bit number, after subscriber access system, according to the channel gain of self and the average channel gain of system, adaptively selected bit number feeds back CDI.

According to the first scheme of the present invention, a kind of limited feedback method has been proposed, comprising: carry out channel estimation process, obtain current channel condition information; According to the average channel information of base station broadcast and the current channel condition information that obtains, select to be used for the number of bits of feedback B of feedback CDI _kAccording to predetermined codeword matching criterion, before code book

The code word that in individual code word, search is mated most; And with the code word sequence number of the code word of coupling that searches, as CDI, feed back to the base station.

Preferably, described limited feedback method can also comprise: after obtaining current channel condition information, select number of bits of feedback before, the number of bits of feedback that is used for feedback channel quality information Channel Quality Information (CQI) with initial setting up, to the current CQI of base station feedback, wherein the average channel information broadcasted of base station is to determine according to the up-to-date CQI that all subscriber equipmenies in system feed back.

Perhaps, the average channel information broadcasted of base station can be that the base station is predetermined.

Preferably, described limited feedback method can also comprise: selecting number of bits of feedback B _kAfterwards, with selected number of bits of feedback B _k, to the current CDI of base station feedback, the while is with the number of bits of feedback that is used for feedback CQI of initial setting up, to base station feedback CQI.

Preferably, can select number of bits of feedback B according to the ratio of current channel gain and average channel gain _kMore preferably, the ratio of current channel gain and average channel gain is larger, selected number of bits of feedback B _kAlso larger.

According to alternative plan of the present invention, a kind of subscriber equipment has been proposed, comprising: receiving element, for the average channel information that receives base station broadcast; Channel estimating unit is used for carrying out channel estimation process, obtains current channel condition information; The number of bits of feedback selected cell is used for selecting to be used for the number of bits of feedback B of feedback CDI according to the average channel information of base station broadcast and the current channel condition information that obtains _kThe codeword matching unit is used for according to predetermined codeword matching criterion, before code book

The code word that in individual code word, search is mated most; And feedback unit, the code word sequence number for the code word of mating most that will search as CDI, feeds back to the base station.

Preferably, before described feedback unit can also be used for after described channel estimating unit obtains current channel condition information, described number of bits of feedback selected cell selects number of bits of feedback, the number of bits of feedback that is used for feedback CQI with initial setting up, to the current CQI of base station feedback, wherein the average channel information broadcasted of base station is to determine according to the up-to-date CQI that all subscriber equipmenies in system feed back.

Perhaps, the average channel information broadcasted of base station can be that the base station is predetermined.

Preferably, described feedback unit can also be used for the selected number of bits of feedback B of described number of bits of feedback selected cell _k, to the current CDI of base station feedback, the while is with the number of bits of feedback that is used for feedback CQI of initial setting up, to base station feedback CQI.

Preferably, described number of bits of feedback selected cell can be selected number of bits of feedback B according to the ratio of current channel gain and average channel gain _kMore preferably, the ratio of current channel gain and average channel gain is larger, the selected number of bits of feedback B of described number of bits of feedback selected cell _kAlso larger.

The invention has the advantages that:

1, on the basis that does not increase system's internal feedback expense, by the adaptive feedback of user CDI, the throughput of final effectively raising system;

2, user's self adaptation Limited Feedback algorithm complex is lower, when adopting optimum adaptive strategy, only needs to carry out the water filling bit in less than M user and selects;

3, when adopting the suboptimum Bit Allocation in Discrete, only need a product calculation.

Do not change the overall structure of original system due to the self adaptation feedback strategy, so the present invention can be merged mutually with existing space division multiple access (SDMA) technology well.

Description of drawings

By below in conjunction with description of drawings the preferred embodiments of the present invention, will make of the present invention above-mentioned and other objects, features and advantages are clearer, wherein:

Fig. 1 is the schematic block diagram according to subscriber equipment 1000 of the present invention.

Fig. 2 is the flow chart based on the self adaptation limited feedback method of mimo system.

Fig. 3 is base station 4 transmit antennas, 1 antenna of user's receiving terminal, and SNR P is 10db, user's code book N=2 ¹⁴, average feedback bits is respectively 6 and at 8 o'clock, and throughput of system is with the graph of relation between number of users.

Fig. 4 is base station 4 transmit antennas, 1 antenna of user's receiving terminal, and in system, number of users is 100, and average feedback bits is respectively 6 and at 8 o'clock, and throughput of system is with the graph of relation of base station transmitting power SNR P.

Fig. 5 is base station 4 transmit antennas, 2 antennas of user's receiving terminal, and SNR P is 10db, user's code book N=2 ¹⁴, average feedback bits is respectively 6 and at 8 o'clock, and throughput of system is with the graph of relation between number of users.

Fig. 6 is base station 4 transmit antennas, 2 antennas of user's receiving terminal, and in system, number of users is 200, and average feedback bits is respectively 6 and at 8 o'clock, and throughput of system is with the graph of relation of base station transmitting power SNR P.

Embodiment

For clear detailed elaboration performing step of the present invention, the below has provided some specific embodiments of the invention, is applicable to the mobile communication system of many antennas, especially multiuser MIMO quadrature space division multiplexing (OSDM) mobile communication system.Need to prove, the invention is not restricted to these and use, but applicable to more other relevant communication systems.

To a preferred embodiment of the present invention will be described in detail, having omitted in the description process is unnecessary details and function for the present invention with reference to the accompanying drawings, obscures to prevent that the understanding of the present invention from causing.

Fig. 1 shows the schematic block diagram according to subscriber equipment 1000 of the present invention.

As shown in Figure 1, subscriber equipment 1000 according to the present invention can comprise: receiving element 1010 is used for receiving the average channel information ‖ H ‖ that broadcast base station 2000; Channel estimating unit 1020 is used for carrying out channel estimation process, obtains current channel condition information H _kNumber of bits of feedback selected cell 1030 is used for according to the average channel information ‖ H ‖ of base station broadcast and the current channel condition information H that obtains _k, select the number of bits of feedback B for the feedback channel directional information _kCodeword matching unit 1040 is used for according to predetermined codeword matching criterion, before code book

The code word that in individual code word, search is mated most

And feedback unit 1050, be used for the code word of mating most that will search

The code word sequence number, as CDI, feed back to base station 2000.

In addition, before feedback unit 1050 also is used for after channel estimating unit 1020 obtains current channel condition information, number of bits of feedback selected cell 1030 selects number of bits of feedback, with the number of bits of feedback that is used for feedback CQI of initial setting up, the 2000 current CQI of feedback to the base station.Perhaps, feedback unit 1050 also is used at the selected number of bits of feedback B of number of bits of feedback selected cell 1030 _kAfter, simultaneously to the base station the current CDI of 2000 feedbacks and CQI information (CDI uses B _kIndividual feedback bits, CQI uses the feedback bits of initial setting up).

Fig. 2 shows the flow chart according to the self adaptation limited feedback method based on mimo system of the present invention.

As shown in Figure 2, in step S200, carry out initialization operation, initial parameter is set in mimo system: the code book C of system, average number of bits of feedback B _av, the self adaptation feedback strategy;

In step S210, the channel estimating unit 1020 of subscriber equipment 1000 is carried out channel estimating, thereby obtains channel condition information H _k

In step S220, the number of bits of feedback selected cell 1030 of subscriber equipment 1000 is according to the channel condition information H of self _kWith the average channel gain information ‖ H ‖ of system, adaptively selected number of bits of feedback B _k, k represents user's sequence number;

In step S230, the codeword matching unit 1040 of subscriber equipment 1000 is according to predetermined codeword matching criterion, at code book

Before

The code word that in individual code word, search is mated most

And

In step S240, the feedback unit 1050 of subscriber equipment 1000 is with code word

The code word sequence number as CDI, feed back to base station 2000.

After base station 2000 receives the code word sequence number, search out its corresponding matrix in self code book, as user's CDI, carry out according to CQI and CDI that the user selects and pre-coding matrix designs, and transmitting data is to subscriber equipment 1000.

Suppose in step S200, the number of transmit antennas of base station 2000 is M, and the reception antenna number of subscriber equipment 1000 is N, and total transmitting power of base station 2000 is P.In system, base station 2000 and subscriber equipment 1000 have identical code book

Code word c in code book _iIt is 1 * Metzler matrix.The average number of bits of feedback of setting subscriber equipment 1000 is B _av, satisfy B _av≤ B _Base, B _BaseIt is user's maximum feedback bit number.

In step S210, the channel estimating unit 1020 of subscriber equipment 1000 is carried out channel estimating, obtains channel condition information H _k=[h _{K, 1}, h _{K, 2}... h _{K, N}], 1≤k≤K, wherein K is intrasystem total number of users, h _{K, i}It is 1 * Metzler matrix.Here, suppose that the channel condition information that subscriber equipment 1000 obtains is complete and accurate.

If the initial connecting system stage, the number of bits of feedback of all subscriber equipmenies 1000 is identical, is B _av

After subscriber equipment 1000 connecting systems, the adaptively selected number of bits of feedback of number of bits of feedback selected cell 1030 is B _k=f (‖ H _k‖, ‖ H ‖), wherein ‖ H ‖ is the average channel gain of system, f (x, y) is about x, the function of y (step S220).

The self adaptation Limited Feedback requires subscriber equipment 1000 to know the average information ‖ H ‖ of system, and the acquisition of ‖ H ‖ has various ways.In the following description, provided two kinds of methods, but the present invention is not limited thereto.

(1) statistics initialization system average channel information

When laying network, the residential quarter is traveled through test, the average channel information ‖ H ‖ that obtains adding up.‖ H ‖ is as a parameter, at the initial access phase of user by base station broadcast to the user.When the user changes serving BS, again be broadcast to the user by new base station.

(2) CQI and CDI divide out feedback

After subscriber equipment 1000 obtained the channel condition information of self, elder generation is 2000 feedback CQI to the base station, and average channel information ‖ H ‖ is calculated according to the CQI of all user feedbacks in base station 2000, then is broadcast to subscriber equipment 1000.Subscriber equipment 1000 obtains feeding back CDI according to the self adaptation feedback strategy again after ‖ H ‖.2000 feedbacks are during CQI to the base station at subscriber equipment 100, and the number of bits of feedback of CQI is by system's initial setting.At the initial access phase of subscriber equipment 1000, the number of bits of feedback of CDI is fixed, and after subscriber equipment 1000 connecting systems, according to the ‖ H ‖ of base station 2000 broadcasting and the channel condition information of self, selects the number of bits of feedback B of CDI _k

Method (1) feasibility is higher, but it is comparatively coarse to obtain ‖ H ‖, can't accurately react the instantaneous channel quality of channel; Method (2) ‖ H ‖ value is comparatively accurate, but need to divide out feedback with CDI and CQI, can cause certain time-delay, and in the telex network stage, requires the base station to continue broadcasting ‖ H ‖ to the user.

In addition, top method (2) has provided a kind of feedback (number of bits of feedback selected cell 1030 selection number of bits of feedback B on opportunity of CQI _k(between step S210 and S220) before).About the feedback opportunity of CQI, can be decided in its sole discretion by the designer, for example, also can select number of bits of feedback B at number of bits of feedback selected cell 1030 _k(identical or different opportunity with the feedback of CDI) afterwards (after step S220, and not restriction of the ordinal relation between step S230 and S240), with the feedback bits of initial setting, the 2000 current CQI of feedback to the base station.

In addition, the present invention provides B _k=f (‖ H _k‖, ‖ H ‖) a kind of suboptimum self adaptation Limited Feedback scheme:

$B_k=\left\{\begin{array}{cc}{B}_{\mathrm{adp}},& 0\&le;B_{\mathrm{adp}}\&le;B_{\mathrm{base}}\\ 0,& B_{\mathrm{adp}}<0\\ B_{\mathrm{base}},& B_{\mathrm{adp}}>B_{\mathrm{base}}\end{array}\right.---\left(1\right)$

Wherein:

Wherein Expression is carried out round to x.

In step S230, when user side has many antennas, need to carry out antenna and merge or select to disturb between the elimination user antenna, that is:

$h_k^{\mathrm{eff}}={\mathrm{\&gamma;}}_k{H}_k---\left(3\right)$

γ in formula (3) _k=[γ ₁, γ ₂... γ _N], ‖ γ _k‖=1, the weights that the expression antenna merges.When user side is single antenna,

The present invention adopts the QBC scheme in Niha Jindal article " Antenna Combining for the MIMODownlink Channel " (referring to " Wireless Communications; IEEETransactions on Volume 7; Issue 10; October 2008Page (s): 3834-3844. "), that is:

${\mathrm{\&gamma;}}_k=\frac{{\left(H_k^{*}{H}_k\right)}^{-1}{H}_k^{*}{\left(s_k^{\mathrm{proj}}\right)}^T}{\left|\right|{\left(H_k^{*}{H}_k\right)}^{-1}{H}_k^{*}{\left(s_k^{\mathrm{proj}}\right)}^T\left|\right|}---\left(4\right)$

In formula (4)

q ₁, q ₂... q _NBe H _kThe N that constructs 1 * M base vector,

Be CDI information, satisfy:

${\hat{h}}_k^{\mathrm{eff}}=\arg \underset{c={\mathrm{<figure-callout\; id="1"\; label="c"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">c</figure-callout>}}_1,{\mathrm{<figure-callout\; id="2"\; label="c"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">c</figure-callout>}}_2,\&CenterDot;\&CenterDot;\&CenterDot;,{\mathrm{<figure-callout\; id="2"\; label="c\; B\; k"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">c</figure-callout>}}_{B_k}}{\max }\left|\right|Q_k^{*}{c}^T\left|\right|---\left(5\right)$

In addition, CQI feedback { g (H _k) two kinds of form: SINR are generally arranged _kWith ‖ H _kThe ‖ feedback.Owing to taking SINR _kFeedback can obtain higher throughput of system, so the present invention only considers SINR _kThe situation of feedback.Order

SINR _kThe form of feedback is as follows:

${\mathrm{SINR}}_k=\frac{\mathrm{\&rho;}{\left|\right|h_k^{\mathrm{eff}}\left|\right|}^2{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2{\mathrm{\&theta;}}_{\mathrm{<figure-callout\; id="1"\; label="k"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">k</figure-callout>}}}{1+\mathrm{\&rho;}{\left|\right|h_k^{\mathrm{eff}}\left|\right|}^2{\mathrm{<figure-callout\; id="2"\; label="sin"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">sin</figure-callout>}}^2{\mathrm{\&theta;}}_k}---\left(6\right)$

ρ=P/M in formula,

Quantization error for channel.

Base station 2000 reconstructs in the code book of self after the code word sequence number that obtains subscriber equipment 1000 feedbacks Rebuilding the CDI of gained and the error between actual CDI can be expressed as:

${\mathrm{<figure-callout\; id="2"\; label="sin"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">sin</figure-callout>}}^2{\mathrm{\&theta;}}_k={\sin }^2(\&angle;({\tilde{h}}_k^{\mathrm{eff}},{\hat{h}}_k^{\mathrm{eff}}))---\left(7\right)$

It can be SUS method (referring to the article " Multi-Antenna Downlink Channels with Limited Feedbackand User Selection " of above-mentioned TaesangYoo) that the performed user in base station 2000 selects, but realization of the present invention is not limited to the method.

Make π represent final selecteed M user.Search for the first time, initial user set K ₀={ 1,2 ..., K} is at K ₀The user of middle searching SINR maximum:

$\mathrm{\&pi;}\left(1\right)=\arg \underset{k\&Element;K_0}{\max }\left({\mathrm{SINR}}_k\right)---\left(8\right)$

After searching out i user, (i+1) individual user's selection principle is:

$\mathrm{\&pi;}(i+1)=\arg \underset{k\&Element;K_i}{\max }\left({\mathrm{SINR}}_k\right)---\left(9\right)$

$K_i=\{1\&le;k\&le;K:|{\hat{h}}_k^{\mathrm{eff}}{\left({\hat{h}}_{\mathrm{\&pi;}\left(j\right)}^{\mathrm{eff}}\right)}^{*}|\&le;\mathrm{\&epsiv;},1\&le;j\&le;i\}---\left(10\right)$

In formula, ε represents to search for the semi-orthogonal upper limit that allows, and general value is 0.25～0.4.

The performed pre-coding matrix design in base station 2000 can be ZFBF method (referring to the article " Multi-Antenna Downlink Channels with LimitedFeedback and User Selection " of above-mentioned Taesang Yoo), but realization of the present invention is not limited to the method.

Select to have obtained u user through the user, u is the element number in set π.Order

Pre-coding matrix is:

W(u)＝H(u) ^*(H(u)H(u) ^*) ^-1 (11)

The beam forming that user π (k) is corresponding is the k row of W, is designated as w _k

Suppose that transmission data corresponding to subscriber equipment k are s _k, base station 2000 place's antennas transmit for

Below, self adaptation Limited Feedback technology is carried out more detailed schematic illustration.

In a multi-user MIMO system, the number of transmit antennas of base station is M, and user's reception antenna number is N, and total transmitting power of base station is P.System's code book is

Code word in code book is 1 * M type matrix.When user's code book produced at random, the article " Antenna Combining for the MIMO Downlink Channel " according to NihaJindal can obtain the relation between user feedback bit number and CDI feedback error:

$E\left({\mathrm{<figure-callout\; id="2"\; label="sin"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">sin</figure-callout>}}^2{\mathrm{\&theta;}}_k\right)=2^{-\frac{B_k}{M-N}}{\left(\begin{array}{c}M-1\\ N-1\end{array}\right)}^{-\frac{1}{M-N}}---\left(12\right)$

When system adopts beam forming technique, user side SINR satisfy:

${\mathrm{SINR}}_k=\frac{\mathrm{\&rho;}{\left|h_k^{\mathrm{eff}}{w}_k\right|}^2}{1+\mathrm{\&rho;}\underset{j\&NotEqual;k}{\mathrm{\&Sigma;}}{\left|h_k^{\mathrm{eff}}{w}_j\right|}^2},1\&le;k\&le;M---\left(13\right)$

ρ=P/M in following formula.Select through the user, the user is quadrature or semi-orthogonal each other, and user SINR is satisfied as can be known according to Taesang Yoo article " Multi-Antenna Downlink Channels with LimitedFeedback and User Selection ":

${\mathrm{SINR}}_k=\frac{\mathrm{\&rho;}{\left|\right|h_k^{\mathrm{eff}}\left|\right|}^2(1-{\mathrm{<figure-callout\; id="2"\; label="sin"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">sin</figure-callout>}}^2{\mathrm{\&theta;}}_k)}{1+\mathrm{\&rho;}{\left|\right|h_k^{\mathrm{eff}}\left|\right|}^2{\mathrm{<figure-callout\; id="2"\; label="sin"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">sin</figure-callout>}}^2{\mathrm{\&theta;}}_k}---\left(14\right)$

Therefore the throughput of final system can be expressed as:

$C=\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}{\log }_2(1+{\mathrm{SINR}}_k)---\left(15\right)$

Can get according to (14) (15) two formulas:

$E\left(C\right)=E\left(\underset{i=k}{\overset{M}{\mathrm{\&Sigma;}}}{\log }_2(1+{\mathrm{SINR}}_k)\right)$

$=E\left(\underset{i=k}{\overset{M}{\mathrm{\&Sigma;}}}{\log }_2\left(\frac{1+\mathrm{\&rho;}{\left|\right|h_k^{\mathrm{eff}}\left|\right|}^2}{1+\mathrm{\&rho;}{\left|\right|h_k^{\mathrm{eff}}\left|\right|}^2{\mathrm{<figure-callout\; id="2"\; label="sin"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">sin</figure-callout>}}^2{\mathrm{\&theta;}}_k}\right)\right)---\left(16\right)$

According to the article " Antenna Combining for the MIMO DownlinkChannel " of Niha Jindal ‖ h as can be known _k ^eff‖ and ‖ H _k‖ satisfies following relation:

${\left|\right|h_k^{\mathrm{eff}}\left|\right|}^2={\left|\right|H_k\left|\right|}^2\mathrm{\&beta;}(M-N+1,N-1)---\left(17\right)$

Formula (14) and formula (17) are brought in (16), can be got by Jensen inequality abbreviation:

$E\left(C\right)\&GreaterEqual;\underset{k=1}{\overset{M}{\mathrm{\&Sigma;}}}{\log }_2\left(\frac{1+\mathrm{\&rho;}{\left|\right|h_k^{\mathrm{eff}}\left|\right|}^2}{1+\mathrm{\&rho;}{\left|\right|H_k\left|\right|}^2\frac{M-N+1}{M}{2}^{-\frac{B_k}{M-N}}{\left(\begin{array}{c}M-1\\ N-1\end{array}\right)}^{-\frac{1}{M-N}}}\right)$

$=\left(\underset{k=1}{\overset{M}{\mathrm{\&Sigma;}}}{\log }_2(1+\mathrm{\&rho;}{\left|\right|H_k\left|\right|}^2\frac{M-N+1}{M})\right)---\left(18\right)$

$-\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{\log }_2(1+\mathrm{\&rho;}{\left|\right|H_k\left|\right|}^2\frac{M-N+1}{M}{2}^{-\frac{B_k}{M-N}}{\left(\begin{array}{c}M-1\\ N-1\end{array}\right)}^{-\frac{1}{M-N}})\right)$

Order:

$E\left(\mathrm{\&Delta;C}\right)=\underset{k=1}{\overset{M}{\mathrm{\&Sigma;}}}{\log }_2(1+{\mathrm{\&rho;}}^{\&prime;}{\left|\right|H_k\left|\right|}^2{2}^{-\frac{B_k}{M-N}}{\left(\begin{array}{c}M-1\\ N-1\end{array}\right)}^{-\frac{1}{M-N}})---\left(19\right)$

In formula (19)

Therefore under the constant condition of the total number of bits of feedback of keeping system, the problem of maximized system capacity can be expressed as:

minΔC

$\mathrm{<figure-callout\; id="1"\; label="c"\; filenames="H2009102621443E00061.png"\; state="\{\{state\}\}">c</figure-callout>}1.\underset{k=1}{\overset{M}{\mathrm{\&Sigma;}}}{B}_k={\mathrm{MB}}_{\mathrm{av}}---\left(20\right)$

c2.0≤B _k≤B _base

C1 is the restriction to the total feedback bits expense of whole system, and C2 shows that the code word sequence number of each user feedback can not surpass the codebook size of self maximum.Find the solution optimum self adaptation Limited Feedback strategy according to Lagrangian method:

$L=\underset{k=1}{\overset{M}{\mathrm{\&Sigma;}}}{\log }_2(1+{\mathrm{\&rho;}}^{\&prime;}{\left|\right|H_k\left|\right|}^2{2}^{-\frac{B_k}{M-N}}{\left(\begin{array}{c}M-1\\ N-1\end{array}\right)}^{-\frac{1}{M-N}})+\mathrm{\&mu;}(\underset{k=1}{\overset{M}{\mathrm{\&Sigma;}}}{B}_k-B_{\mathrm{Tol}})---\left(21\right)$

Order $\frac{\&PartialD;L}{\&PartialD;B_k}=0,$ Have:

$B_k=\left\{\begin{array}{cc}{B}_{\mathrm{adp}},& 0\&le;B_{\mathrm{adp}}\&le;B_{\mathrm{base}}\\ 0,& B_{\mathrm{adp}}<0\\ B_{\mathrm{base}},& B_{\mathrm{adp}}>B_{\mathrm{base}}\end{array}\right.---\left(1\right)$

B _kAdaptively selectedly be similar to water filling and distribute, it should be noted that number of bits of feedback must be integer, so B _kIn the water filling assigning process, adopt round to carry out.With B _kBring into

Obtain the horizontal λ of water filling, and then obtain each user's self adaptation number of bits of feedback.

It is to be based upon selecteed user to know on each other channel condition information basis that optimum self adaptation Limited Feedback is derived, and knows the final selecteed user of system before feeding back channel state information, is applicable to system user and counts in the less situation of K.

When the user can't know the final selecteed user of system, utilize the statistical average channel gain log of base station broadcast ₂(‖ H ‖ ²) be similar in λ

The number of users that M ' distributes for participating in iteration.With log ₂(‖ H ‖ ²) approximate can getting in substitution formula (22):

Consider that user in real system does not generally know other users' channel condition information, so the present invention relates generally to this suboptimal self adaptation Limited Feedback scheme.

Suppose the number of transmit antennas M=4 of base station, user's reception antenna number is N={1,2}, system's code book

B _Base=14, namely user's maximum feedback bit number is 14.The average number of bits of feedback B of user _av=6,8}, the average SNR P of transmitting terminal are 0～25db, and number of users 10～500, Channel Modeling is the rayleigh distributed channel.

Fig. 3 is base station 4 transmit antennas, 1 antenna of user's receiving terminal, and SNR P is 10db, user's code book N=2 ¹⁴, average feedback bits is respectively 6 and at 8 o'clock, and throughput of system is with the graph of relation between number of users.

Following six kinds of different situations have been compared in emulation:

1) throughput of system when user's unity feedback (perfect CSIT), the feedback bits of this moment levels off to infinity;

2) strategy of the ZFBF when user's fixed bit Limited Feedback 6 (Fix ZFBF (B=6));

3) average bit is fed back to 6, and the user takes ZFBF strategy of the present invention (AdaptiveZFBF (B=6));

4) strategy of the ZFBF when user's fixed bit Limited Feedback 8 (Fix ZFBF (B=8));

5) average bit is fed back to 8, and the user takes ZFBF strategy of the present invention (AdaptiveZFBF (B=8));

6) accidental beam shaping strategy (RBF).

Can find out, along with the increase of number of users, the throughput of system under Different Strategies all increases to some extent, and takes the solution of the present invention, can obtain higher throughput of system than traditional fixed bit Limited Feedback.

Fig. 4 is base station 4 transmit antennas, 1 antenna of user's receiving terminal, and in system, number of users is 100, and average feedback bits is respectively 6 and at 8 o'clock, and throughput of system is with the graph of relation of base station transmitting power SNR P.

Emulation 2000 times, compared following six kinds of different situations:

1) throughput of system when user's unity feedback (perfect CSIT), the feedback bits of this moment levels off to infinity;

2) strategy of the ZFBF when user's fixed bit Limited Feedback 6 (Fix ZFBF (B=6));

3) average bit is counter is 6, and the user takes the ZFBF strategy (AdaptiveZFBF (B=6)) of name of the present invention;

4) strategy of the ZFBF when user's fixed bit Limited Feedback 8 (Fix ZFBF (B=8));

5) average bit is fed back to 8, and the user takes ZFBF strategy of the present invention (AdaptiveZFBF (B=8));

6) accidental beam shaping strategy (RBF).

Can find out, along with the increase of SNR P, the throughput of system under Different Strategies all increases to some extent, and takes the solution of the present invention, can obtain higher throughput of system than traditional fixed bit Limited Feedback.

Fig. 5 is base station 4 transmit antennas, 2 antennas of user's receiving terminal, and SNR P is 10db, user's codebook size N=2 ¹⁴, average feedback bits is respectively 6 and at 8 o'clock, and throughput of system is with the graph of relation between number of users.

Following six kinds of different situations have been compared in emulation:

1) throughput of system when user's unity feedback (perfect CSIT (N=1)) is only considered the situation of user side single antenna herein;

2) strategy of the ZFBF when user's fixed bit Limited Feedback 6 (Fix ZFBF (B=6));

3) average bit is fed back to 6, and the user takes ZFBF strategy of the present invention (AdaptiveZFBF (B=6));

4) strategy of the ZFBF when user's fixed bit Limited Feedback 8 (Fix ZFBF (B=8));

5) average bit is fed back to 8, and the user takes ZFBF strategy of the present invention (AdaptiveZFBF (B=8));

6) accidental beam shaping strategy (RBF (N=1)) is only considered the situation of user side single antenna herein.

Can find out, for the situation of the many antennas of user side, take the solution of the present invention, can obtain higher throughput of system than traditional fixed bit Limited Feedback.

Fig. 6 is base station 4 transmit antennas, 2 antennas of user's receiving terminal, and in system, number of users is 200, and average feedback bits is respectively 6 and at 8 o'clock, and throughput of system is with the graph of relation of base station transmitting power SNR P.

Following six kinds of different situations have been compared in emulation:

1) throughput of system when user's unity feedback (perfect CSIT (N=1)) is only considered the situation of user side single antenna herein;

2) strategy of the ZFBF when user's fixed bit Limited Feedback 6 (Fix ZFBF (B=6));

3) average bit is fed back to 6, and the user takes ZFBF strategy of the present invention (AdaptiveZFBF (B=6));

4) strategy of the ZFBF when user's fixed bit Limited Feedback 8 (Fix ZFBF (B=8));

5) average bit is fed back to 8, and the user takes ZFBF strategy of the present invention (AdaptiveZFBF (B=8));

6) accidental beam shaping strategy (RBF (N=1)) is only considered the situation of user side single antenna herein.

Can find out, for the situation of the many antennas of user side, take the solution of the present invention, can obtain higher throughput of system than traditional fixed bit Limited Feedback.

In order further to illustrate the present invention when not increasing overhead, can improve throughput of system, table 1 has provided when number of users K=10～500, and SNR P is 10db, the fixed bit Limited Feedback is 6 o'clock, the overhead of the present invention and fixed bit Limited Feedback.Because this programme is identical with the CQI feedback of fixed bit Limited Feedback scheme, thereby only compared the feedback overhead of two schemes CDI.Data shown in table 1 are the statistical value after analog simulation 2000 times, and the unit of CDI feedback overhead is bit, and the unit of throughput is bps/Hz.

Table 1

M=4, N=2, P=10db, B _av=6, B _Base=14 o'clock, the CDI expense of system

Number of users K	The present invention program B ad	Fixed bit feedback B ft	Expense difference B ad-B ft	Throughput difference R ad-R ft
					10	59.136	60	-0.864	0.0578
50	295.372	300	-4.628	0.3524
					100	590.728	600	-9.272	0.4591
200	1181.42	1200	-18.58	0.6132
					500	2953.42	3000	-46.58	0.9012

As can be seen from Table 1, take adaptive Limited Feedback not only not increase the feedback overhead of CDI, can save number of bits of feedback on the contrary.

Table 2 has provided when number of users K=10～500, and SNR P is 10db, and the fixed bit Limited Feedback is 8 o'clock, the overhead of the present invention and fixed bit Limited Feedback.Because this programme is identical with the CQI feedback of fixed bit Limited Feedback scheme, thereby only compared the feedback overhead of two schemes CDI.Data shown in table 2 are the statistical value after analog simulation 2000 times, and the unit of CDI feedback overhead is bit, and the unit of throughput is bps/Hz.

Table 2

M=4, N=2, P=10db, B _av=8, B _Base=14 o'clock, the CDI expense of system

Number of users K	The present invention program B ad	Fixed bit feedback B ft	Expense difference B ad-B ft	Throughput difference R ad-R ft
					10	79.161	80	-0.839	0.0473
50	395.308	400	-4.692	0.2100
					100	790.716	800	-9.284	0.3206
200	1581.23	1600	-18.77	0.3982
					500	3953.36	4000	-46.62	0.5417

As can be seen from Table 2, take adaptive bit feedback can save number of bits of feedback equally, and the elevator system throughput.

By above detailed description, the present invention uses the mode of self adaptation Limited Feedback, can when reducing the CDI of system expense, effectively increase throughput of system.Do not change the overall structure of system due to the self adaptation Limited Feedback, therefore can be good at compatible existing orthogonal beams forming technique, have very high using value.

So far invention has been described in conjunction with the preferred embodiments.Should be appreciated that, those skilled in the art can carry out various other change, replacement and interpolations without departing from the spirit and scope of the present invention.Therefore, scope of the present invention is not limited to above-mentioned specific embodiment, and should be limited by claims.

Claims (14)

1. subscriber equipment comprises:

Receiving element is for the average channel information that receives base station broadcast;

Channel estimating unit is used for carrying out channel estimation process, obtains current channel condition information;

The number of bits of feedback selected cell is used for selecting to be used for the number of bits of feedback B of feedback channel directional information according to the average channel information of base station broadcast and the current channel condition information that obtains _k

The codeword matching unit is used for according to predetermined codeword matching criterion, before code book

The code word that in individual code word, search is mated most; And

Feedback unit, the code word sequence number for the code word of mating most that will search as the channel directional information, feeds back to the base station.

2. subscriber equipment according to claim 1 is characterized in that:

Before described feedback unit also is used for after described channel estimating unit obtains current channel condition information, described number of bits of feedback selected cell selects number of bits of feedback, the number of bits of feedback that is used for feedback channel quality information with initial setting up, to the current channel quality information of base station feedback

Wherein the average channel information broadcasted of base station is to determine according to the up-to-date channel quality information that all subscriber equipmenies in system feed back.

3. subscriber equipment according to claim 1 is characterized in that:

The average channel information that broadcast the base station is that the base station is predetermined.

4. subscriber equipment according to claim 1 is characterized in that:

Described feedback unit also is used for the selected number of bits of feedback B of described number of bits of feedback selected cell _k, to the current channel directional information of base station feedback, the while is with the number of bits of feedback that is used for feedback channel quality information of initial setting up, to the base station feedback channel quality information.

5. subscriber equipment according to claim 3 is characterized in that:

Described feedback unit also is used for the selected number of bits of feedback B of described number of bits of feedback selected cell _k, to the current channel directional information of base station feedback, the while is with the number of bits of feedback that is used for feedback channel quality information of initial setting up, to the base station feedback channel quality information.

6. one of according to claim 1～5 described subscriber equipmenies is characterized in that:

Described number of bits of feedback selected cell is selected number of bits of feedback B according to the ratio of current channel gain and average channel gain _k

7. subscriber equipment according to claim 6 is characterized in that:

The ratio of current channel gain and average channel gain is larger, the selected number of bits of feedback B of described number of bits of feedback selected cell _kAlso larger.

8. limited feedback method comprises:

Carry out channel estimation process, obtain current channel condition information;

According to the average channel information of base station broadcast and the current channel condition information that obtains, select to be used for the number of bits of feedback B of feedback channel directional information _k

According to predetermined codeword matching criterion, before code book

The code word that in individual code word, search is mated most; And

The code word sequence number of the code word of coupling with searching as the channel directional information, feeds back to the base station.

9. limited feedback method according to claim 8 also comprises:

After obtaining current channel condition information, select number of bits of feedback before, with the number of bits of feedback that is used for feedback channel quality information of initial setting up, to the current channel quality information of base station feedback,

Wherein the average channel information broadcasted of base station is to determine according to the up-to-date channel quality information that all subscriber equipmenies in system feed back.

10. limited feedback method according to claim 8 is characterized in that:

The average channel information that broadcast the base station is that the base station is predetermined.

11. limited feedback method according to claim 8 also comprises:

Selecting number of bits of feedback B _kAfterwards, with selected number of bits of feedback B _k, to the current channel directional information of base station feedback, the while is with the number of bits of feedback that is used for feedback channel quality information of initial setting up, to the base station feedback channel quality information.

12. limited feedback method according to claim 10 also comprises:

Selecting number of bits of feedback B _kAfterwards, with selected number of bits of feedback B _k, to the current channel directional information of base station feedback, the while is with the number of bits of feedback that is used for feedback channel quality information of initial setting up, to the base station feedback channel quality information.

13. one of according to claim 8～12 described limited feedback methods is characterized in that:

Select number of bits of feedback B according to the ratio of current channel gain and average channel gain _k

14. limited feedback method according to claim 13 is characterized in that:

The ratio of current channel gain and average channel gain is larger, selected number of bits of feedback B _kAlso larger.

Images