CN102013906B - Multi-mode transmission method based on mailmonitor normalization beamforming and rate control - Google Patents

Abstract

The invention discloses a multi-mode transmission method based on per-user normalized beamforming and rate control, which is characterized in that in a system using per-user normalized beamforming and rate control, the user feeds back quantization quality indications to the base station, The base station reconstructs the user's quantization error through the quantization quality indicator, so that the user's signal-to-interference-noise ratio in different transmission modes can be estimated more accurately during scheduling, so as to select the optimal transmission mode. The present invention effectively overcomes the problem of limited system and capacity interference in the actual application scene of the traditional per-user normalized beamforming and rate control algorithm, and is more suitable for use in the actual system.

Description

Multimode transmission method based on per-user normalized beamforming and rate control

technical field

The invention belongs to the technical field of transmission methods for wireless communication multi-antenna multi-user systems, in particular to a multi-antenna multi-user system adopting a per-user normalized beamforming and rate control (Per-User Unitary Beamforming and Rate Control, PU ² RC) transmission strategy multimode transmission method.

Background technique

"On the optimality of multiantenna broadcast scheduling using zero-forcing beamforming" Selected Areas in Communications, IEEE Journalon., vol.24, no.3, PP.528-541, Mar.2006 ) pointed out that when the complete channel information (CSIT) is known by the base station, the multi-user multiple-input multiple-output (MU-MIMO) technology using space-division multiplexing can make the system obtain complete space-division multiplexing gain, but in practice In the system, especially in the frequency division duplex system, because the bandwidth of the uplink feedback link is limited, an infinite number of bits is required to fully quantize the user channel and the bit information needs to be fed back to the base station through the uplink feedback link, so the base station has known The assumption of perfect channel information (CSIT) is difficult to achieve.

"Journal of Electronic Information and Communication Society"("On the Performance of Multiuser MIMO Systems inWCDMA/HSDPA: Beamforming, Feedback and User Diversity," IEICE Trans.Commun., vol.E89-B, no.8, pp.2161-2169 , Aug.2006) introduced a multi-user MIMO precoding and user selection scheme based on limited feedback: per-user normalized beamforming and rate control (Per-User Unitary Beamforming and Rate Control, PU ² RC). When the number of system users K tends to infinity, this scheme can achieve the same spatial multiplexing gain as the case where the sender knows the perfect channel information (CSIT), but when the number of actual system users is relatively small, the multiuser diversity gain (multiuser diversity gain ) is reduced, this scheme has a large loss in system performance.

"Multi-Antenna Downlink Channels with Limited Feedback and User Selection," Selected Areas in Communications, IEEE Journal on, vol.25, no.7, p.1478-1491, Jul.2007 ) pointed out that when the receiving end adopts Channel Vector Quantization (Channel Vector Quantization, CVQ) method to feed back channel information to the sending end, if the number of bits used to quantize the channel is constant, then, as the system signal-to-noise ratio (SNR) increases , the sum rate of the multi-user MIMO system will be limited due to the increase of interference between users.

"Achievable throughput of multi-mode multiuser MIMO with imperfect CSI constraints," Information Theory, 2009.ISIT2009.IEEE International Symposium on) pointed out that in a limited feedback system, the sender can Overcome traditional multi-user multi-antenna system (MU-MIMO) system and capacity interference at high signal-to-noise ratio (SNR) through multi-mode multi-user multi-antenna transmission (Multi-Mode MU-MIMO) with variable number of users per service The downside of being limited is that it is an analysis based on statistical information.

Contents of the invention

The purpose of the present invention is to propose a multi-mode transmission method based on per-user normalized beamforming and rate control, so as to realize the multi-mode transmission in which the number of users served by the base station at the same time changes adaptively with the system environment, and improve the system performance. and capacity, overcome the shortcomings of existing methods that are limited by interference at high signal-to-noise ratios, supplement the shortcomings of existing user selection algorithms, and make them more suitable for practical system applications.

The present invention is based on the multi-mode transmission method of per-user normalized beamforming and rate control (PU ² RC), characterized in that: in the downlink multi-antenna multi-user system adopting per-user normalized beamforming and rate control transmission strategy , assuming that there is one base station and multiple users in the cell, the base station is equipped with two to eight transmitting antennas, and the user end is equipped with one receiving antenna; each user feeds back the precoding matrix index (Precoding Matrix Index, PMI) to the base station At the same time, the user's channel quantization quality indicator is fed back to the base station through 1-bit information. The base station uses this 1-bit information to reconstruct the quantization error according to the principle of scalar quantization, and uses the quantization error to accurately estimate the user's Signal-to-interference-noise ratio, so as to select a set of sending users suitable for the current system environment to realize multi-mode transmission;

The specific operation steps are as follows:

Part 1: User Feedback

为正交集即

m＝1，...M为单位向量且两两正交，M为基站端发送天线数；对于第k个用户，其在码本集合W中找出与自身信道矢量方向最接近的码本，并将该码本在码本集合中的索引作为向基站反馈的预编码矩阵索引(PMI)，即PMI＝j^**M+i^*其中

h为基站到用户之间的信道；用户向基站反馈预编码矩阵索引(PMI)所需的比特数B＝log2(M*G)；Step A. The user feeds back the precoding matrix index (PMI) to the base station: in a system using per-user normalized beamforming and rate control (PU ² RC) transmission strategy, its codebook set is W={W ¹ , ..., W ^G }, where

is an orthogonal set that

m=1,... M is a unit vector and is orthogonal to each other, and M is the number of transmitting antennas at the base station; for the kth user, it finds the codebook closest to its own channel vector direction in the codebook set W , and use the index of the codebook in the codebook set as the precoding matrix index (PMI) fed back to the base station, that is, PMI=j ^* *M+i ^* where

h is the channel between the base station and the user; the user feeds back the required bit number B=log2(M*G) of the precoding matrix index (PMI) to the base station;

为该用户向基站反馈的预编码矩阵索引所对应的码本，h为该用户的实际信道；用户将信道量化误差sin²(∠(w_PMI，h))与做比较，式中M为基站端的发送天线数，B为预编码矩阵索引的量化比特数；如果

则量化质量较差；反之量化质量较好，用户用1比特将信道量化质量反馈给基站：用户反馈信道量化质量为1时表示量化质量较好，用户反馈信道量化质量为0时表示量化质量较差；Step B. Quantization quality feedback: the quantization error between the codebook corresponding to the precoding matrix index fed back by each user to the base station and its actual channel is defined as sin ² (∠(w _PMI , h)), where

is the codebook corresponding to the precoding matrix index fed back by the user to the base station, h is the actual channel of the user; the user uses the channel quantization error sin ² (∠(w _PMI , h)) and For comparison, M in the formula is the number of transmitting antennas at the base station, and B is the number of quantized bits of the precoding matrix index; if

The quantization quality is poor; otherwise, the quantization quality is better, and the user uses 1 bit to feed back the channel quantization quality to the base station: when the user feedback channel quantization quality is 1, it means that the quantization quality is good, and when the user feedback channel quantization quality is 0, it means that the quantization quality is relatively low. Difference;

Step C. The user feeds back the channel gain ||h|| ² as user channel quality information (Channel Quality Indicator, CQI) to the base station so that the base station can perform scheduling effectively;

Part II: Base Station User Scheduling

Step D. Restore the quantization error of each user: for each user, if the channel quantization quality of its feedback is better, restore its quantization error as If the channel quantization quality is poor, restore its quantization error as

Step E. Precoding matrix selection, that is, selecting the precoding matrix W ⁱ used for this transmission from W={W ¹ ,...,W ^G }:

找出选择以

为预编码矩阵索引的所有用户，并根据信号干扰噪声比粗略估算公式

计算出用户的粗略信号干扰噪声比SINR^coarse，其中P为噪声归一化后的基站总发送功率；在所有选择预编码向量

为预编码矩阵索引的用户中选出粗略信号干扰噪声比SINR^coarse最大的用户记为

并记其粗略信号干扰噪声比SINR^coarse为

如果对于预编码向量

没有用户选择其作为自己的预编码矩阵索引，则记该预编码向量对应的用户为空集

且其粗略信号干扰噪声比

对所有预编码矩阵找出使得系统和容量最大的预编码矩阵

并记使用预编码向量

i＝1，2，...，M作为预编码矩阵索引PMI中粗略信号干扰噪声比SINR^coarse最大的用户集合为

For each precoding vector

find options to

All users indexed for the precoding matrix, and roughly estimated according to the signal-to-interference-noise ratio formula

Calculate the user's rough signal-to-interference-noise ratio SINR ^coarse , where P is the total transmit power of the base station after noise normalization; in all selected precoding vectors

Select the user with the largest rough signal-to-interference-noise ratio SINR ^coarse among the users indexed for the precoding matrix as

And record its rough signal-to-interference-noise ratio SINR ^coarse as

If for the precoded vector

If no user chooses it as its own precoding matrix index, the user corresponding to the precoding vector is recorded as an empty set

and its rough signal-to-interference-to-noise ratio

Find the precoding matrix that maximizes the system and capacity for all precoding matrices

And remember to use the precoded vector

i=1, 2,..., M is used as the user set with the largest rough SINR coarse SINR ^coarse in the precoding matrix index PMI is

初始化系统估算和容量R(A′)＝0；对第E步中所得到的用户集合A中的用户按照粗略信号干扰噪声比SINR^coarse从大到小排序，并记排序结果为U＝{u₁，..，u_M}；在信号干扰噪声比粗略估算中，设基站同时服务M个用户，当基站同时服务N≤M个用户时，利用重构后的量化误差得到信号干扰噪声比的精细估算公式：Step F. User selection process 1: record the user selection result as A′, and initialize the user selection result as an empty set

Initialize the system estimation and capacity R(A')=0; sort the users in the user set A obtained in step E according to the rough signal-to-interference-noise ratio SINR ^coarse from large to small, and record the sorting result as U={u ₁ ,..., u _M }; in the rough estimation of SINR, assume that the base station serves M users at the same time, and when the base station serves N≤M users at the same time, the SINR can be obtained by using the quantization error after reconstruction Fine estimate formula:

${\mathrm{<span\; class="notranslate">\; SINR</span>}}^{\mathrm{<span\; class="notranslate">\; precise</span>}}<span\; class="notranslate">\; =</span>\frac{\frac{\mathrm{<span\; class="notranslate">\; P</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="cos"\; filenames="HDA0000036013600000011.png"\; state="\{\{state\}\}">cos</figure-callout></span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; \&theta;</span>}}{\frac{\mathrm{<span\; class="notranslate">\; P</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}\frac{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="sin"\; filenames="HDA0000036013600000011.png"\; state="\{\{state\}\}">sin</figure-callout></span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; \&theta;</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Based on the finely estimated signal-to-interference-noise ratio SINR ^precise , the steps of user selection process 1 are as follows:

F.1. Initialize the iteration variable i=1;

F.2. Judging whether the i-th user u _i in the user set U is an empty set, if yes, as soon as the user selection process ends, go to step F.5; otherwise, go to step F.3;

进行计算；F.3. Calculate the system sum capacity R(A′∪u _i ) when the i-th user u _i joins the user selection result A′, and compare it with the system sum when the i-th user u _i does not join the user selection result A′ capacity R(A′), if R(A′)≤R(A′∪u _i ), then skip to step F.4; otherwise, once the user selection process is over, skip to step F.5; where the user selects the result The sum capacity R(A') of A' uses the fine sum capacity calculation formula

Calculation;

F.4. Update the user selection result A'=A'∪u _i and the iteration variable i=i+1; compare the iteration variable i with the number M of transmitting antennas at the base station, and skip to F.2 if i≤M; Otherwise, as soon as the user selection process ends, skip to F.5;

F.5. If the user selection result A' is equal to the user set A, skip to step G; otherwise, the user scheduling process ends, and the user selection result A' is the user scheduling result;

中未被第F步执行后用户选择结果A′中用户使用的预编码向量在未被选择到的用户集合中选择与该预编码向量配对最好的用户，并将其加入到用户选择结果A′中；当用户向基站反馈的预编码矩阵索引(PMI)为而实际传输时采用预编码向量

时，其精细的信号干扰噪声比估算公式为：Step G. User selection process 2: For the precoding matrix used in this transmission

The precoding vector used by the user in the user selection result A′ after step F is not executed Select the best user paired with the precoding vector from the unselected user set, and add it to the user selection result A'; when the precoding matrix index (PMI) fed back by the user to the base station is In actual transmission, the precoding vector is used

When , its fine signal-to-interference-to-noise ratio estimation formula is:

${\mathrm{<span\; class="notranslate">\; SINR</span>}}^{\mathrm{<span\; class="notranslate">\; precise</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; SINR</span>}}^{\mathrm{<span\; class="notranslate">\; coarse</span>}}{<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; max</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; g</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}^{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{\mathrm{<span\; class="notranslate">\; SINR</span>}}^{\mathrm{<span\; class="notranslate">\; coarse</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; Z</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&NotEqual;</span>\mathrm{<span\; class="notranslate">\; L</span>}}{<span\; class="notranslate">\; \{</span><span\; class="notranslate">\; |</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; g</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}<span\; class="notranslate">\; \}</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

The set of precoding vectors actually used in transmission is Z, and the minimum angle between adjacent vectors in the codebook set W is α. In the above formula

The specific selection process steps are as follows:

G.1. Initialize the iteration variable i=1;

是否已被使用，如果预编码向量

已被使用，令用户k为空集：

跳往步骤G.5；如果预编码向量

未被使用则跳往步骤G.3；G.2. Judging the precoding vector

has been used, if the precoding vector

has been used, let user k be the empty set:

Skip to step G.5; if the precoding vector

If not used, skip to step G.3;

G.3. Use the fine signal-to-interference-noise ratio estimation formula (1) to traverse the unselected user set, find out the user that makes the fine signal-to-interference-to-noise ratio the largest, and denote it as k;

进行计算，其中对于第F步选中的用户，即用户反馈的预编码矩阵索引(PMI)与其传输时使用的预编码向量相等的用户，其精细的信号干扰噪声比用信号干扰噪声比的精细估算公式(1)计算，对于第G步选中的用户，即用户反馈的预编码矩阵索引(PMI)与其传输时使用的预编码向量不相等的用户，其精细的信号干扰噪声比用精细的信号干扰噪声比估算公式(2)计算；G.4. Calculate the system and capacity R(A′∪k) when the user k with the largest fine SINR joins the user selection result, and the system when the user k with the largest fine SINR does not join the user selection result Compared with the capacity R(A′), if R(A′)≤R(A′∪k), skip to step G.5; otherwise, the user scheduling process ends, and the user selection result A′ is the scheduling result; the user The sum capacity R(A') of the selection result A' uses the fine sum capacity calculation formula

For the user selected in step F, that is, the user whose precoding matrix index (PMI) fed back by the user is equal to the precoding vector used in transmission, the fine signal-to-interference-noise ratio is estimated using the fine-grained signal-to-interference-noise ratio Formula (1) calculates that for the user selected in step G, that is, the user whose precoding matrix index (PMI) fed back by the user is not equal to the precoding vector used in transmission, the fine signal-to-interference-noise ratio is equal to the fine signal-to-interference-noise ratio Noise ratio estimation formula (2) calculation;

G.5. Update the user selection result and the iteration variable A'=A'∪k, i=i+1, compare the iteration variable i with the number of base station transmitting antennas M, if i≤M, then skip to step G.2; Otherwise, the user scheduling process ends, and the user selection result A' is the scheduling result.

The present invention is a technical solution for realizing multi-mode transmission by using instantaneous channel information, wherein the number of users served simultaneously varies from 1 to the number of transmitting antennas at the base station. When the existing per-user normalized beamforming and rate control technology is applied to an actual system, because the number of users K in the actual system is about ten, which is relatively small, and the performance of per-user normalized beamforming and rate control is progressive The optimality is based on the fact that the number of system users tends to be infinite, and the multi-user diversity gain decreases when the number of users is small; On the other hand, in the case of high signal-to-noise ratio, the interference of the system is limited due to the excessive interference between users, which also causes a large loss of the system and capacity. In the method of the present invention, in a system based on per-user normalized beamforming and rate control, the user feeds back the quantization quality indication to the base station, and the base station reconstructs the quantization error of the user through the quantization quality indication, so that more accurate estimation can be made during scheduling The signal-to-interference-noise ratio of users in different transmission modes is calculated to select the optimal transmission mode. The present invention effectively overcomes the problem of limited system and capacity interference in the actual application scene of the traditional per-user normalized beamforming and rate control algorithm, and is more suitable for use in the actual system.

Description of drawings:

Figure 1 shows the system using the multi-mode transmission method based on per-user normalized beamforming and rate control when the number of system users K=10, the number of base station antennas M=4, and the number of system quantization bits B=4, and other transmission strategies A schematic diagram of the system comparison.

Figure 2 shows the system using the multi-mode transmission method based on per-user normalized beamforming and rate control when the number of system users K=10, the number of base station antennas M=4, and the number of system quantization bits B=10, and other transmission strategies A schematic diagram of the system comparison.

Detailed ways:

Embodiments of the method are described below in conjunction with the accompanying drawings.

Example 1:

Assuming that this embodiment adopts a downlink multi-antenna multi-user system with 10 users, the number of base station antennas is M=4, the number of receiving end antennas is 1, and the system codebook W={W ¹ ,...,W _G } is randomly generated produced in a manner. Assuming that each user can accurately obtain its own channel state information (CSIR), h _j is the channel vector from the base station to user j, and each element is a complex Gaussian random variable that obeys independent and identical distribution, with a mean of 0 and a variance of 1 .

The specific steps of the multimode transmission method based on per-user normalized beamforming and rate control (PU ² RC) in this embodiment are as follows:

Part 1: User Feedback

为正交集即

m＝1，..M为单位向量且两两正交，M为基站端发送天线数；对于第k个用户，其在码本集合W中找出与自身信道矢量方向最接近的码本，并将该码本在码本集合中的索引作为向基站反馈的预编码矩阵索引，即PMI＝j^**M+i^*其中

h为基站到用户之间的信道；用户向基站反馈预编码矩阵索引PMI所需的比特数B＝log2(M*G)；Step A. Feedback the precoding matrix index (PMI) to the base station: in the system adopting per-user normalized beamforming and rate control (PU ² RC) transmission strategy, its codebook set is W={W ¹ ,. .., W ^G }, where

is an orthogonal set that

m=1, ..M is a unit vector and is orthogonal to each other, and M is the number of transmitting antennas at the base station; for the kth user, it finds the codebook closest to its own channel vector direction in the codebook set W, And the index of the codebook in the codebook set is used as the precoding matrix index fed back to the base station, that is, PMI=j ^* *M+i ^* where

h is the channel between the base station and the user; the user feeds back the required bit number B=log2(M*G) of the precoding matrix index PMI to the base station;

为该用户向基站反馈的预编码矩阵索引所对应的码本，h为该用户的实际信道；用户将信道量化误差

做比较，其中M为基站端的发送天线数，B为预编码矩阵索引的量化比特数；如果

则量化质量较差；反之量化质量较好，用户用1比特将信道量化质量反馈给基站：用户反馈信道量化质量为1表示量化质量较好，用户反馈信道量化质量为0表示量化质量较差；Step B. Quantization quality feedback: the quantization error between the codebook corresponding to the precoding matrix index that each user feeds back to the base station and its actual channel is defined as sin ² (∠(w _PMI , h)), where

is the codebook corresponding to the precoding matrix index fed back by the user to the base station, h is the actual channel of the user; the user quantizes the channel quantization error

For comparison, M is the number of transmit antennas at the base station, and B is the number of quantized bits of the precoding matrix index; if

Then the quantization quality is poor; otherwise, the quantization quality is better, and the user uses 1 bit to feed back the channel quantization quality to the base station: the user feedback channel quantization quality is 1, which means the quantization quality is good, and the user feedback channel quantization quality is 0, which means the quantization quality is poor;

Step C. The user feeds back the channel gain ||h|| ² as user channel quality information (CQI) to the base station so that the base station can perform scheduling effectively;

Part II: Base Station User Scheduling

如果信道量化质量为差时，还原其量化误差为

Step D. Restore the quantization error of each user: For each user, if the channel quantization quality of its feedback is good, restore its quantization error as

If the channel quantization quality is poor, restore its quantization error as

Step E. Precoding matrix selection: select the precoding matrix W ⁱ used for this transmission from W={W ¹ ,...,W ^G }:

找出选择以为预编码矩阵索引的所有用户，并根据信号干扰噪声比粗略估算公式

计算出用户的粗略信号干扰噪声比SINR^coarse，其中P为噪声归一化后的基站总发送功率；在所有选择预编码向量

为预编码矩阵索引的用户中选出粗略信号干扰噪声比SINR^coarse最大的用户记为

并记其粗略信号干扰噪声比SINR^coarse为

如果对于预编码向量

没有用户选择其作为自己的预编码矩阵索引，则记该预编码向量对应的用户为空集

且其粗略信号干扰噪声比为0(即

)；对所有预编码矩阵找出使得系统和容量最大的预编码矩阵

并记使用

作为预编码矩阵索引中粗略信号干扰噪声比SINR^coarse最大的用户集合为用户集合

For each precoding vector

find options to All users indexed for the precoding matrix, and roughly estimated according to the signal-to-interference-noise ratio formula

Calculate the user's rough signal-to-interference-noise ratio SINR ^coarse , where P is the total transmit power of the base station after noise normalization; in all selected precoding vectors

Select the user with the largest rough signal-to-interference-noise ratio SINR ^coarse among the users indexed for the precoding matrix as

And record its rough signal-to-interference-noise ratio SINR ^coarse as

If for the precoded vector

If no user chooses it as its own precoding matrix index, the user corresponding to the precoding vector is recorded as an empty set

And its rough signal-to-interference-noise ratio is 0 (ie

); Find the precoding matrix that maximizes the system and capacity for all precoding matrices

And remember to use

As the user set with the largest rough signal-to-interference-noise ratio SINR ^coarse in the precoding matrix index is the user set

)，初始化系统估算和容量为0(即R(A′)＝0)；对第E步中所得到的用户集合A中的用户按照粗略信号干扰噪声比SINR^coarse从大到小排序，并记排序结果为U＝{u₁，..，u_M}，在信号干扰噪声比SINR粗略估算中，假设基站同时服务M个用户，考虑当基站同时服务N≤M个用户时，利用重构后的量化误差可以得到信号干扰噪声比SINR的精细估算公式(1)即

基于精细的信号干扰噪声比SINR^precise，用户选择过程一的步骤如下：Step F. User selection process one: record the user selection result of the base station as A ', and initialize the user selection result as an empty set (i.e.

), initialize the system estimation and capacity to 0 (that is, R(A′)=0); sort the users in the user set A obtained in step E according to the rough signal-to-interference-noise ratio SINR ^coarse from large to small, and record The sorting result is U={u ₁ ,..,u _M }. In the rough estimation of SINR, it is assumed that the base station serves M users at the same time. Considering that when the base station serves N≤M users at the same time, the reconstructed The quantization error of the signal-to-interference-noise-noise ratio SINR can be obtained from the fine estimation formula (1) that is

Based on the fine signal-to-interference-noise ratio SINR ^precise , the steps of user selection process 1 are as follows:

F.1. Initialize the iteration variable i=1;

F.2. Judging whether the i-th user u _i in the user set U is an empty set, if yes, as soon as the user selection process ends, go to step F.5; otherwise, go to step F.3;

F.3. Calculate the system and capacity R(A′∪u _i ) when user u _i joins the user selection result A′, and compare it with the system and capacity R(A′) when u _i does not join the user selection result A′ If R(A′)≤R(A′∪u _i ), go to step F.4; otherwise, as soon as the user selection process ends, go to step F.5; where the sum capacity R( A') Use Calculation;

F.4. Update the user selection result A'=A'∪u _i and the iteration variable i=i+1; compare the iteration variable i with the number of base station transmitting antennas M, if i≤M, then skip to step F.2; otherwise Once the user selection process is over, skip to step F.5

F.5. If the user selection result A' is equal to the user set A, skip to step G; otherwise, the user scheduling process ends, and the user selection result A' is the user scheduling result.

中未被第F步执行后用户选择结果A′中用户使用的预编码向量

在未被选择到的用户集合中选择与该预编码向量配对最好的用户，并将其加入到用户选择结果A′中；当用户向基站反馈的预编码矩阵索引(PMI)为而实际传输时采用预编码向量

时，其精细的信号干扰噪声比可以用公式(2)即进行估算，其中传输中实际使用到的预编码向量集合为Z，码本集合W中相邻向量的最小夹角为α，上式中

Step G. User selection process 2: For the precoding matrix used in this transmission

The precoding vector used by the user in the user selection result A′ after step F is not executed

Select the best user paired with the precoding vector from the unselected user set, and add it to the user selection result A'; when the precoding matrix index (PMI) fed back by the user to the base station is In actual transmission, the precoding vector is used

When , its fine signal-to-interference-to-noise ratio can be expressed by formula (2): Estimated, where the set of precoding vectors actually used in transmission is Z, and the minimum angle between adjacent vectors in the codebook set W is α, in the above formula

The specific selection process is as follows:

G.1. Initialize the iteration variable i=1;

是否已被使用，如果预编码向量

已被使用，令用户

跳往步骤G.5；如果预编码向量

未被使用则跳往步骤G.3；G.2. Judging the precoding vector

has been used, if the precoding vector

has been used, making the user

Skip to step G.5; if the precoding vector

If not used, skip to step G.3;

G.3. Use the fine signal-to-interference-noise ratio estimation formula (1) to traverse the unselected user set, find out the user that makes the fine signal-to-interference-to-noise ratio the largest, and denote it as k;

进行计算，其中对于第F步选中的用户，即用户反馈的预编码矩阵索引(PMI)与其传输时使用的预编码向量相等的用户，其精细的信号干扰噪声比用(1)式计算，对于第G步选中的用户，即用户反馈的预编码矩阵索引(PMI)与其传输时使用的预编码向量不相等的用户，其精细的信号干扰噪声比用(2)式计算；G.4. Calculate the system and capacity R(A′∪k) when the user k with the largest fine SINR joins the user selection result, and the system when the user k with the largest fine SINR does not join the user selection result Compared with the capacity R(A′), if R(A′)≤R(A′∪k), skip to step G.5; otherwise, the user scheduling process ends, and the user selection result A′ is the scheduling result; the user The sum capacity R(A') of the selection result A' utilizes

Calculate, wherein for the user selected in step F, that is, the user whose precoding matrix index (PMI) fed back by the user is equal to the precoding vector used in transmission, its fine signal-to-interference-noise ratio is calculated by formula (1), for The user selected in step G, that is, the user whose precoding matrix index (PMI) fed back by the user is not equal to the precoding vector used in transmission, has a fine signal-to-interference-noise ratio calculated by formula (2);

G.5. Update the user selection result A'=A'∪k and the iteration variable i=i+1, compare the iteration variable i and the number of transmitting antennas M at the base station, if i≤M, then skip to step G.2; otherwise, the user When the scheduling process ends, the user selection result A' is the scheduling result.

Accompanying drawing 1 shows when the number of system users K=10, the number of base station antennas M=4, and the number of system quantization bits B=4, the system using the multimode transmission method based on normalized beamforming and rate control per user and the system using other transmission methods Schematic diagram of the system comparison of strategies; Figure 2 is a multi-mode transmission method based on per-user normalized beamforming and rate control when the number of system users K=10, the number of base station antennas M=4, and the number of system quantization bits B=10 Schematic diagram of the comparison between the system and systems using other transmission strategies; where P represents the total transmission power of the system after noise normalization.

It can be seen from Figure 1 and Figure 2 that when the total number of feedback bits B=4 and 10, under various transmission strategies, the total transmission power P of the system after noise normalization is from -5dB to 30dB in the downlink multi-antenna multi-user system system and capacity. Wherein A adopts the system and capacity of multi-mode transmission based on PU ² RC proposed by the present invention; B represents the system and capacity of adopting the traditional PU ² RC transmission strategy; C represents the use of the PU ² RC transmission strategy that the base station always sends with full flow ( That is, the base station serves M users at the same time) system and capacity. For the system adopting the traditional PU ² RC transmission strategy and the system adopting the PU ² RC transmission strategy transmitted by the base station in full flow, the B feedback bits are used as the precoding matrix index feedback. For the PU ² RC based on the present invention For multi-mode transmission, (B-1) of the B feedback bits is used to feed back the precoding matrix index, and one is used to feed back the quantization quality of the user.

The curves in Figure 1 and Figure 2 show the following characteristics:

1. When the number of feedback bits used is B=4, the system and capacity of the traditional PU ² RC and full-flow PU ² RC methods are limited by interference as the total transmission power of the system increases, while the system based on PU ^{2 RC} The RC multi-mode transmission method overcomes this shortcoming. The system and capacity increase with the increase of ^the total transmission power ^. Compared with the PU ² RC that is sent at full flow, the system and capacity have a gain of 37.3% and 62.0% respectively; when the total transmission power of the system is small, although the multimode transmission method based on PU ² RC uses more PMI feedback The number of bits is less, but compared with the traditional PU ² RC and the PU ² RC sent in full flow, there is no obvious loss in performance.

2. When the number of feedback bits used is B=10, the performance of the three schemes is improved due to the increase of the quantization precision, while the multimode transmission based on PU ² RC still maintains the same performance as other schemes in the entire transmission power range. Advantages of both options.

The present invention aims at overcoming the limitation of system and capacity interference when the transmission power increases under limited feedback, and proposes this multi-mode transmission scheme based on PU ² RC. When the total number of bits fed back by users is constant, each user feeds back At the same time as its own PMI, 1 bit is used to feed back the quantization quality of the channel, so that the base station can more accurately calculate the estimated signal-to-interference-noise ratio of each user in different transmission modes, so that it can be better in various transmission modes. to choose from. Compared with the traditional PU ² RC algorithm, in the system using per-user normalized beamforming and rate control, the user feeds back the quantization quality indication to the base station, and the base station reconstructs the quantization error of the user through the quantization quality indication, Therefore, the signal-to-interference-noise ratio of users in different transmission modes can be estimated more accurately during scheduling, so as to select the optimal transmission mode. The present invention effectively overcomes the problem of limited system and capacity interference in the actual application scene of the traditional per-user normalized beamforming and rate control algorithm, and is more suitable for use in the actual system.

Claims (1)

1. multimode transmission method based on the control of the moulding of every user's normalization beam and speed, it is characterized in that: adopting the moulding of every user's normalization beam and speed to control in the descending multi-antenna multi-user system of transmission policy, if a base station and a plurality of user are arranged in the residential quarter, the base station end is furnished with two to eight transmitting antennas, and user side is furnished with a reception antenna; Each user is in the base station feedback pre-coding matrix index, indicate the amount of information by 1 bit to feed back to the base station user's channel quantitative quality, the base station utilizes this 1 bit information according to the principle reconstruct quantization error of scalar quantization, and carrying out utilizing quantization error accurately to estimate user's Signal Interference and Noise Ratio when the user selects, thereby select the transmission user who adapts with current system environments to gather, realize the multimode transmission;

The concrete operations step is as follows:

First: user feedback

A step. the user is to the base station feedback pre-coding matrix index: in adopting the system of the moulding of every user's normalization beam and speed control transmission policy, its code book is gathered and is

Wherein $W^i=\{w_1^i,...,w_M^i\}$ For orthogonal set is M=1 ... M is unit vector and pairwise orthogonal, and M is base station end number of transmit antennas; For k user, it is gathered at code book In find out and the immediate code book of self channel vector direction, and with the index of this code book in code book set as the pre-coding matrix index to base station feedback, i.e. PMI=j ^** M+i ^*Wherein

H is that the base station is to the channel between the user; The user is to the required bit number B=log2 (M*G) of base station feedback pre-coding matrix index (PMI);

B step. quantize quality feedback: each user is defined as sin to the quantization error of the code book interchannel actual with it of the pre-coding matrix index correspondence of base station feedback ²(∠ (w _PMI, h)), wherein

For this user to the corresponding code book of the pre-coding matrix index of base station feedback, h is this user's actual channel; The user is with channel quantitative error sin ²(∠ (w _PMI, h)) with

Compare, M is the number of transmit antennas of base station end in the formula, and B is the quantizing bit number of pre-coding matrix index; If

Then quantize second-rate; Otherwise it is better to quantize quality, and the user gives the base station with 1 bit with the channel quantitative quality feedback: user feedback channel quantitative quality is that 1 expression quantification quality is better, and user feedback channel quantitative quality is that 0 expression quantizes second-rate;

The C step. the user is with channel gain || h|| ²Give the base station so that effectively dispatch the base station as the user channel quality feedback information;

Second portion: base station user scheduling

D step. reduce each user's quantization error: to each user, if when the channel quantitative quality of its feedback is better, reduces its quantization error and be

When if the channel quantitative quality is relatively poor, reduces its quantization error and be ${\sin }^2\mathrm{\&theta;}=\frac{2M-1}{2M}{2}^{-\frac{N}{M-1}};$

E step. the selection of pre-coding matrix, namely from

The middle pre-coding matrix W that selects this transmission to use ⁱ:

For each precoding vector

Find out selection with

Be all users of pre-coding matrix index, and according to the rough estimation equation of Signal Interference and Noise Ratio

Calculate user's rough Signal Interference and Noise Ratio SINR ^Coasre, wherein P is the total transmitted power in base station after the noise normalization; Select precoding vector at all

Select rough Signal Interference and Noise Ratio SINR among the user for pre-coding matrix index ^CoarseMaximum user is designated as

And remember its rough Signal Interference and Noise Ratio SINR ^CoarseBe SINR _i ^jIf for precoding vector

Do not have the user to select it as the pre-coding matrix index of oneself, remember that then the user of this precoding vector correspondence is empty set

And its rough Signal Interference and Noise Ratio

All pre-coding matrixes are found out make the pre-coding matrix of system and capacity maximum $j^{*}=\underset{j=1,...,G}{\arg \max }\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\log 2(1+{\mathrm{SINR}}_i^j),$ And note is used precoding vector $w_i^{j^{*}},i=\mathrm{1,2},...,M$ As rough Signal Interference and Noise Ratio SINR in the pre-coding matrix index ^CoarseMaximum user's set is

The F step. user's selection course one: note user selection result is

Initialization user selection result is empty set Initialization system estimation and capacity

To E resulting user's set in the step

In the user according to rough Signal Interference and Noise Ratio SINR ^CoarseOrdering from big to small, and the note ranking results is

In the rough estimation of Signal Interference and Noise Ratio, establish the base station and serve M user simultaneously, when the base station is served N≤M user simultaneously, utilize quantization error after the reconstruct to obtain the meticulous estimation equation of Signal Interference and Noise Ratio

${\mathrm{SINR}}^{\mathrm{precise}}=\frac{\frac{P}{N}{\left|\right|h\left|\right|}^2{\cos }^2\mathrm{\&theta;}}{\frac{P}{N}\frac{N-1}{M-1}{\left|\right|h\left|\right|}^2{\sin }^2\mathrm{\&theta;}+1},$

Signal Interference and Noise Ratio SINR based on meticulous estimation ^Precise, the step of user's selection course one is as follows:

F.1, initialization iteration variable i=1;

F.2, judge user's set

In i user u _iWhether be empty set, if then user's selection course one finishes, jump toward step F .5; Otherwise jump toward step F .3;

F.3, calculate i user u _iAdd the access customer selection result

The time system and capacity

And with i user u _iDo not add the access customer selection result The time system and capacity

Compare, if Then jump toward step F .4; Otherwise user's selection course one finishes, and jumps toward step F .5; User's selection result wherein

And capacity

Utilize meticulous and calculation of capacity formula Calculate;

F.4, upgrade user's selection result

And iteration variable i=i+1; Iteration variable i and base station end number of transmit antennas M are compared, if i≤M then jump toward F.2; Otherwise user's selection course one finishes, and jumping is past F.5;

If user's selection result F.5

Gather with the user

Equate, then jump the step toward G; Otherwise user's scheduling process finishes, user's selection result

Be user's scheduling result;

The G step. user's selection course two: for the pre-coding matrix that uses in this transmission In carried out back user's selection result by F step The precoding vector that middle user uses

In the user's set that is not chosen to, select the user best with this precoding vector pairing, and it is joined user's selection result

In; When the user to the pre-coding matrix index of base station feedback is

And adopt precoding vector during actual transmissions

The time, its meticulous Signal Interference and Noise Ratio estimation equation is:

Wherein the actual precoding vector set that uses is in the transmission

The code book set The minimum angle of middle adjacent vector is α, in the following formula $\mathrm{\&epsiv;}=\sqrt{2(1-\cos \frac{\mathrm{\&alpha;}}{2});}$

Concrete selection course step is as follows:

G.1, initialization iteration variable i=1;

G.2, judge precoding vector

Whether be used, if precoding vector Be used, make that user k is empty set:

Jump toward step G.5; If precoding vector Be not used and then jump toward step G.3;

G.3, utilize meticulous Signal Interference and Noise Ratio estimation equation to travel through non-selected user set, find out and make the user of meticulous Signal Interference and Noise Ratio maximum be designated as k;

System and capacity when G.4, the user k of the meticulous Signal Interference and Noise Ratio maximum of calculating adds the access customer selection result

And system and capacity when not adding the access customer selection result with the user k of meticulous Signal Interference and Noise Ratio maximum

Compare, if

Then jump toward step G.5; Otherwise user's scheduling process finishes, user's selection result

Be scheduling result; User's selection result

And capacity

Utilize meticulous and calculation of capacity formula Calculate, wherein go on foot the user who chooses for F, be the user that precoding vector that the pre-coding matrix index of user feedback uses during with its transmission equates, its meticulous Signal Interference and Noise Ratio calculates with the meticulous Signal Interference and Noise Ratio estimation equation of F in the step, go on foot the user who chooses for G, the unequal user of precoding vector who uses when being the pre-coding matrix index of user feedback and its transmission, its meticulous Signal Interference and Noise Ratio calculates with the meticulous Signal Interference and Noise Ratio estimation equation of G in the step;

G.5, upgrade user's selection result and iteration variable

I=i+1 compares iteration variable i and base station end number of transmit antennas M, if toward step G.2 i≤M then jumps; Otherwise user's scheduling process finishes, user's selection result

Be scheduling result.

Images