CN102256301A - User selection method for simultaneously meeting unicast and multicast services - Google Patents

Abstract

The invention provides a user selection method for simultaneously meeting unicast and multicast services, which comprises the following steps of: (1) overlapping unicast on the original a carrier wave distributed to multicast, and establishing a user selection model; (2) ensuring that the speed of the multicast service on the carrier wave can not be reduced through block diagonalization (BD) and power control; and (3) selecting a unicast user set with largest unicast and speed. In the method provided by the invention, during multicast service transmission, the demand of a user needing the unicast service while needing the multicast service is considered simultaneously. In an OFDM (Orthogonal Frequency Division Multiplexing) system, unicast is overlapped on the subcarrier wave distributed to the multicast through BD on the premise of ensuring that the receiving property of the poorest multicast user can not be influenced. Speed maximization and no generation of new poorest link user can be ensured by selecting the optimal unicast user set and the corresponding power control scheme.

Description

A kind of user choosing method that satisfies clean culture and multicast service simultaneously

Technical field

The invention belongs to wireless communication technology field, especially a kind of user choosing method that satisfies clean culture and multicast service simultaneously.

Background technology

Along with the development of wireless network, the user is more and more higher to the quality requirement of business, and limited frequency spectrum resources does not more and more satisfy user's requirement.Therefore propose a plurality of users are transmitted the multicasting technology of identical data, can well utilize frequency spectrum, improved efficiency of resource.Yet because a plurality of subscriber channel link condition of multicast service differ, in order to satisfy all users' business demand, the data rate of multicast often is subject to the poorest user of link, and the better user's of link resource does not make full use of like this.Simultaneously, when a plurality of users for example received video and this multicast service of TV programme, they also had the corresponding for example demand of unicast service such as phone, note.In order to address this problem, the also situation of demand clean culture when considering the user's request multicast simultaneously, research transmits clean culture simultaneously on a subcarrier and multicast service has real value.

Summary of the invention

The objective of the invention is to be subject to the poorest user of link, and good user's resource does not fully obtain the problem utilized in order to overcome existing multicast system; Propose a kind of user choosing method that satisfies clean culture and multicast service simultaneously, this method is when guaranteeing that multicast speed does not reduce, and the speed of maximization clean culture effectively raises the utilance of frequency spectrum, thereby improves the performance of whole system.

The objective of the invention is to be achieved through the following technical solutions:

The present invention is a kind of user choosing method that satisfies clean culture and multicast service simultaneously, plays special character and is in the following steps:

1) clean culture that on original carrier wave of distributing to multicast, superposes;

2) guarantee can not reduce multicast service speed on this carrier wave by the control of BD and power;

3) select to guarantee the unicast rate maximum by unicast user.

In technique scheme, the concrete operations of step 1) are as follows:

1-1), the result that arranges according to wireless scene, scheduling multicast service and unicast service on all carrier waves;

1-2), in the carrier wave of scheduling multicast, according to the poorest user performance of link, calculate the speed of multicast;

1-3), in the carrier wave of scheduling multicast, to multi-case data stack unicast data stream,, calculate the speed of multicast according to the poorest user performance of multicast this moment link:

In technique scheme, just unicast stream is together sent by supercomposed coding and multicast.

In technique scheme, before the stack clean culture, the SNR of user i multicast signals is:

${\mathrm{SNR}}_{m,i}^{\left(k\right)}=\frac{P_m{\left|w_{m,k}^H{h}_i^{\left(k\right)}\right|}^2}{N_0}---\left(1\right)$

After the stack, the SINR of user i multicast signals is:

${\mathrm{SINR}}_{m,i}=\frac{P_m{\left|w_{m,k}^H{h}_i^{\left(k\right)}\right|}^2}{P_u\underset{j\∈U_k}{\mathrm{\Σ}}{\left|w_{j,k}^H{h}_i^{\left(k\right)}\right|}^2+N_0}---\left(2\right)$

User j, j ∈ U _k, unicast signal SINR is:

${\mathrm{SINR}}_{u,j}^{\left(k\right)}=\frac{{\left|w_{j,k}^H{h}_j^{\left(k\right)}\right|}^2}{\underset{n\∈U_k,n\≠j}{\mathrm{\Σ}}{\left|w_{n,k}^H{h}_i^{\left(k\right)}\right|}^2+N_0}---\left(3\right)$

In above-mentioned technical scheme, step 2) comprise following operation:

2-1), in the carrier wave of scheduling multicast, according to block diagonalization, make unicast data stream all send at the poorest link user's orthogonal subspaces;

2-2), the poorer user of link occurs because clean culture is disturbed, unicast stream is carried out power control in order to prevent.

In technique scheme, reach the constant purpose of the poorest user rate of multicast link by BD earlier, guarantee not produce other poorer users by power control then, so just can guarantee that multicast speed is constant, unicast data stream can superpose simultaneously.

In technique scheme, step 3) comprises following operation:

3-1), all possible unicast user collection of traversal, according to step 2) calculate various unicast user collection unicast rate down, the clean culture collection of selection clean culture and speed maximum;

3-2), the user of suboptimum collects selection strategy, according to step 2) at first select the maximum user of a speed, selected to select next user on the basis this user then, if speed increase then continue to select, otherwise the user selects to finish.

The present invention has following beneficial effect: the present invention at first by block diagonalization solved unicast data stream to former multicast the poorest user's performance can not exert an influence, guaranteed can not produce by power control then than the worse user of the poorest user performance of former multicast, carry out the user at last and select to have guaranteed under the prerequisite that does not reduce multi-case data speed, maximization unicast data speed; The lower suboptimum user choosing method of complexity has been proposed simultaneously.

Description of drawings

Fig. 1 is system model figure of the present invention;

Transmitter block diagram when Fig. 2 is clean culture and multicast stack;

Receiver block diagram when Fig. 3 is clean culture and multicast stack;

Fig. 4 is the flow chart of the inventive method.

Embodiment

The present invention is described in further detail with concrete embodiment below in conjunction with accompanying drawing.

Core concept of the present invention is in distributing to the carrier wave of multicast, because multicast is subject to the poorest user of link, the user's that link is good resource is not fully utilized, and therefore proposing does not influence under the prerequisite of multicast speed in assurance, by supercomposed coding multicast and clean culture stack is sent.In order to reach the purpose that does not influence multicast speed, at first by block diagonalization (BD, BlockDiagonalization) solve unicast data stream to former multicast the poorest user's performance can not exert an influence, guaranteed can not produce by power control then than the worse user of the poorest user performance of former multicast, guaranteed under the prerequisite that does not reduce multi-case data speed maximization unicast data speed at last by user's selection strategy.The lower suboptimum user selection strategy of complexity has been proposed simultaneously.

The present invention is a kind of user choosing method that satisfies clean culture and multicast service simultaneously, and it comprises:

1) clean culture that on original carrier wave of distributing to multicast, superposes, concrete steps are as follows:

1-1), the ofdm system among the present invention is under a cellular cell scene, the base station of this cellular cell (BS, Base Station) has N antenna, travelling carriage (the MS that has M single antenna in the sub-district, Mobile Station), and M＞N, total K carrier wave can be used to transmit data.Also demand multicast service in the time of each MS demand unicast service.Therefore, total M unicast stream and a multicast stream can be used to transmission.But unicast stream can not all be used for sending, so the user who needs selection to be used for transmitting, as shown in Figure 1.One total K subcarrier can be used to transmit data, when the minimum SNR of all users on k carrier wave greater than threshold value SNR _ThThe time, 1≤k≤K, k subcarrier is used for transmitting multicast service, otherwise the transmission unicast service, as shown in Figure 4.And on the subcarrier of each transmission multicast, the through-put power of multicast is P _m, the through-put power of clean culture is P _uMake h _iThe down channel vector of expression user i, h _iIn each element all be independent same distribution, and to obey average be that 0 variance is 1 multiple Gaussian Profile.We suppose known all users' of BS channel condition information, and at a frame length in the time, channel does not change.

Suppose that k subcarrier allocation given multicast, on k subcarrier, the channel vector of user i is

At first select unicast user collection U _k, satisfy 1≤| U _k|≤M, wherein | U _k| user's number in the expression user set.If the signal of multicast is x, U _{K uses}The signal of family clean culture is y _j, j ∈ U _k, and E[|x| is arranged ²]=E[|y _j| ²]=1, j ∈ U _kTo multicast service, the signal that each antenna transmission is identical, i.e. multicast emission vector is

Wave beam forming vector to j user's unicast signal is w _{J, k}, j ∈ U _k, all w _{J, k}Constituted the wave beam forming vector w of clean culture _{U, k}As shown in Figure 2, with the emission that stacks up of the unicast signal behind multicast and the wave beam forming, so being received as of i user

$r_{i,k}=\sqrt{P_m}{w}_{m,k}^H{h}_i^{\left(k\right)}x+\sqrt{P_u}\underset{j\∈U_k}{\mathrm{\Σ}}{w}_{j,k}^H{h}_i^{\left(k\right)}{y}_i+n_{i,k},n_{i,k}\∈\mathrm{CN}(0,N_0).$

Wherein, n _{I, k}Represent the white Gaussian noise of i user on k subcarrier, N ₀The power spectral density that refers to noise.

At receiving terminal, all users directly translate multi-case data, and user j ∈ U _kAlso need with receiving data r _{J, k}After deducting multi-case data, translate unicast data again, as shown in Figure 3.

1-2), the stack clean culture before, the SNR of user i multicast signals on k subcarrier is:

${\mathrm{SNR}}_{m,i}^{\left(k\right)}=\frac{P_m{\left|w_{m,k}^H{h}_i^{\left(k\right)}\right|}^2}{N_0}---\left(1\right)$

1-3), the stack after, the SINR of user i multicast signals on k subcarrier is:

${\mathrm{SINR}}_{m,i}^{\left(k\right)}=\frac{P_m{\left|w_{m,k}^H{h}_i^{\left(k\right)}\right|}^2}{P_u\underset{j\∈U_k}{\mathrm{\Σ}}{\left|w_{j,k}^H{h}_i^{\left(k\right)}\right|}^2+N_0}---\left(2\right)$

User j, j ∈ U _k, the SINR of unicast signal on k subcarrier is:

${\mathrm{SINR}}_{u,j}^{\left(k\right)}=\frac{{\left|w_{j,k}^H{h}_j^{\left(k\right)}\right|}^2}{\underset{n\∈U_k,n\≠j}{\mathrm{\Σ}}{\left|w_{n,k}^H{h}_i^{\left(k\right)}\right|}^2+N_0}---\left(3\right)$

Target of the present invention is to guarantee to promote the clean culture performance on the basis that multicast speed does not reduce, and the speed of multicast is on k subcarrier $R_m^{\left(k\right)}=\log (1+\underset{i}{\min }\{{\mathrm{SNR}}_{m,i}^{\left(k\right)}\left\}\right),i=\mathrm{1,2},...,M,$ Be multicast speed only and the poorest subscriber-related of link, therefore the poorest user's SNR is constant after the clean culture that superposes as long as guarantee link, just can guarantee that multicast service is unaffected.

In sum, the Optimization Model below proposing: select by the user, power control, and wave beam forming design are guaranteeing under the constant prerequisite of multicast speed, the speed of maximization unicast data and.

$\underset{U_k,P_u,w_{j,k}}{\max }{R}_u^{\left(k\right)}$

$s.t.:\underset{i}{\min }{\mathrm{SINR}}_{m,i}^{\left(k\right)}=\underset{i}{\min }{\mathrm{SNR}}_{m,i}^{\left(k\right)},i=\mathrm{1,2},...,M$ (4)

$R_u^{\left(k\right)}={\mathrm{\Σ}}_{j\∈U_k}{\log }_2(1+{\mathrm{SINR}}_{u,j}^{\left(k\right)}),k=\mathrm{1,2},...,K$

j∈U _k

Wherein first condition of (4) formula is represented unicast signal to not influence of the poorest user of Multicast Channel, and after the unicast signal that promptly superposes, the poorest link user's of multicast performance does not change.Second condition refers to k the total speed of the clean culture on the subcarrier.

2) guarantee can not reduce multicast service speed on this carrier wave by the control of BD and power, concrete steps are as follows:

Because formula (4) is a multivariable optimization problem, a plurality of variablees are separated, the selection branch that is about to design, power control and the user of wave beam forming comes and finds the solution.The at first wave beam forming vector design and power control of research when the user determines; Secondly determine U by the sub-optimal algorithm of the method for exhaustion and low complex degree respectively.Since only consider the problem on the single carrier wave, simple in order to write, " k " ignored do not write below.

2-1), separate by BD and never influence the poorest user performance of link, i.e. w _uDetermine

Guarantee not influence multicast speed, will guarantee that the poorest user of link is not subjected to the interference of clean culture.Adopt BD to solve this problem, the wave beam forming vector of clean culture is obtained from the poorest link user's channel orthogonal subspaces, so just can guarantee that the wave beam of the data of clean culture concentrates on other user's direction, and the poorest user is not subjected to any influence.

Disturb owing to exist between unicast data, also will carry out BD simultaneously, therefore will be BD together, obtain wave beam forming vector w user and the poorest user of link of U to the unicast user of selecting _uSo, because MS is a single antenna, and total N the antenna of BS, so U _kElement number can not surpass N-1.

Can make not influence of the poorest user of former multicast by the wave beam forming vector that obtains with the poorest user BD, but can other user's reception condition be impacted, for the minimum SINR of the multicast after the clean culture that guarantees to superpose and the preceding minimum SNR of stack the same big, just need to guarantee can not produce the new the poorest user of link, this problem can solve by the power control to clean culture.

2-2), guarantee can not produce the poorer user of link, i.e. P by power control _uDetermine

Unicast power control strategy:, therefore introduce the unicast power control strategy because the interference of clean culture may cause multicast to receive the poorest user's change.

For the bonding multicast data does not have influence to multicast transmission speed, the unicast data behind the BD is carried out power control, the power P of distribution _uSeveral conditions below needing to satisfy:

(1), the stack interference that clean culture caused does not influence multicast transmission speed, i.e. the SINR of multicast minimum;

(2), power needs big as far as possible;

(3), power can not surpass a certain upper limit, is designated as P _Max

Therefore in order to guarantee that not having the poorest user of new link is born, unicast power must satisfy

$\underset{i}{\min }\left\{\frac{P_m{\left|w_m^H{h}_i\right|}^2}{P_{u,i}\underset{j\∈U}{\mathrm{\Σ}}{\left|w_j^H{h}_i\right|}^2+N_0}\right\}=\underset{i}{\min }\left\{\frac{P_m{\left|w_m^H{h}_i\right|}^2}{N_0}\right\}=\frac{P_m{\left|w_m^H{h}_{\min }\right|}^2}{N_0}$

Find the solution P _{U, i}, again because unicast power can not surpass P _Max, the gross power of therefore distributing to unicast service is P _u=min{P _{U, i}, P _Max.

So model simplification is a following formula:

$\underset{U}{\max }{R}_u$

$s.t.:\underset{i}{\min }{\mathrm{SINR}}_{m,i}=\underset{i}{\min }{\mathrm{SNR}}_{m,i},k=\mathrm{1,2},...,K$

|U|≤N-1 (5)

$R_u={\mathrm{\Σ}}_{j\∈U}{\log }_2(1+\frac{p_j{\left|w_j{h}_j\right|}^2}{N_0})$

3) select to guarantee the unicast rate maximum that concrete steps are as follows by unicast user:

In above-mentioned optimization problem (5) formula, has only variable of U.The principle that at first clear and definite U selects:

(1), the signal of selected unicast user need concentrate on the good multicast users place of link condition as far as possible to the interference of multicast signals, promptly do not influence the poorest user of link, so the poorest user of link can be not selected;

(2), unicast user requires channel relevancy each other not high among the U, promptly guarantee clean culture behind the BD with the speed maximum.

3-1), because total M of user one, the number of users among the U is no more than N-1, therefore only needs all possible user's collection of traversal, just can obtain the result of optimum.Ergodic algorithm is as follows:

Step 1: at first seek k the poorest user i of the link on the subcarrier, make Ω=1,2 ..., M} i, U=φ;

Step 2: enumerate all possible U from Ω, L altogether,

Be designated as U ^(l), l=1,2 ..., L;

Step 3: every kind of U is carried out BD, and distribute power;

Step 4: the speed of calculating every kind of U $R_u^{\left(l\right)}={\mathrm{\Σ}}_{j\∈U^{\left(l\right)}}{\log }_2(1+\frac{p_j{\left|w_j^H{h}_j\right|}^2}{N_0});$

Step 5: obtain optimum U sequence number, $l_{\mathrm{opt}}=\underset{1\&le;l\&le;L}{\mathrm{<figure-callout\; id="1"\; label="arg\; max"\; filenames="HDA0000076017290000022.png"\; state="\{\{state\}\}">arg</figure-callout>}\max }{R}_u^{\left(l\right)}.$

3-2), because ergodic algorithm complexity height proposes a kind of suboptimum selection algorithm that reduces complexity below:

Step 1: at first seek k the poorest user i of the link on the subcarrier, make Ω=1,2 ..., M} i, U=φ;

Step 2: each user among the Ω carries out power control respectively after user i is BD, selects the user π (1) of speed maximum, then Ω=Ω π (1), U=U+{ π (1) }, t=1;

Step 3: iteration: t=t+1, if t＞N-1, algorithm finishes.Otherwise the user among the Ω seeks respectively and user i, π (1) ..., π (t-1) is BD, and power control is then selected and π (1) ..., π (t-1) and the π speed maximum (t), $R_u^{\left(t\right)}={\mathrm{\&Sigma;}}_{j\&Element;U^{\left(l\right)}}{\log }_2(1+\frac{p_j{\left|w_j^H{h}_j\right|}^2}{N_0})$ For and speed, wherein right

If

Algorithm finishes.Otherwise, Ω=Ω-{ π (t) }, U=U+{ π (t) }.

Above-mentioned solution can obtain three parameters: wave beam forming vector w _u, the power P of clean culture _uAnd the unicast service user collects U, and total flow chart as shown in Figure 4.

Claims (5)

1. a user choosing method that satisfies clean culture and multicast service simultaneously is used for a cellular cell, and this cellular cell comprises a base station with N antenna, the travelling carriage of M single antenna, and M＞N, and total K carrier wave can be used to transmit data; It is characterized in that:

1) user's preference pattern is set up in the clean culture that superposes on original carrier wave of distributing to multicast;

2) guarantee can not reduce multicast service speed on this carrier wave by the control of block diagonalization and power;

3) the unicast user collection of selection clean culture and speed maximum.

2. the user choosing method that satisfies clean culture and multicast service simultaneously according to claim 1 is characterized in that, the concrete grammar of step 1) is as follows:

1-1) before the calculating stack clean culture, the SNR of user i multicast signals on k subcarrier, its computing formula is as follows:

${\mathrm{SNR}}_{m,i}^{\left(k\right)}=\frac{P_m{\left|w_{m,k}^H{h}_i^{\left(k\right)}\right|}^2}{N_0}---\left(1\right)$

Wherein, P _mThe power that refers to multicast on the single sub-carrier,

The wave beam forming vector that refers to multicast on k the subcarrier is represented the signal that each antenna transmission is identical,

The channel coefficients of representing k i user on the subcarrier, N ₀Refer to noise power spectral density;

1-2) after the calculating stack, the SINR of user i multicast signals on k subcarrier, its computing formula is as follows:

${\mathrm{SINR}}_{m,i}^{\left(k\right)}=\frac{P_m{\left|w_{m,k}^H{h}_i^{\left(k\right)}\right|}^2}{P_u\underset{j\&Element;U_k}{\mathrm{\&Sigma;}}{\left|w_{j,k}^H{h}_i^{\left(k\right)}\right|}^2+N_0}---\left(2\right)$

Wherein, P _uThe power that refers to clean culture on the single sub-carrier, w _{J, k}The unicast signal wave beam forming vector that refers to j user on k the subcarrier, U _kRefer to selected unicast user collection on k the subcarrier;

1-3) speed of multicast is before the stack clean culture $R_m^{\left(k\right)}=\log (1+\underset{i}{\min }\{{\mathrm{SNR}}_{m,i}^{\left(k\right)}\left\}\right),i=\mathrm{1,2},...,M;$

Multicast speed before making multicast speed after the stack equal to superpose, its model is as follows:

$\underset{U_k,P_u,w_{j,k}}{\max }{R}_u^{\left(k\right)}$

$s.t.:\underset{i}{\min }{\mathrm{SINR}}_{m,i}^{\left(k\right)}=\underset{i}{\min }{\mathrm{SNR}}_{m,i}^{\left(k\right)},i=\mathrm{1,2},...,M$ (3)

$R_u^{\left(k\right)}={\mathrm{\&Sigma;}}_{j\&Element;U_k}{\log }_2(1+{\mathrm{SINR}}_{u,j}^{\left(k\right)}),k=\mathrm{1,2},...,K$

j∈U _k

Wherein,

Represent U on k the subcarrier _kTotal speed of middle unicast user.

3. the concrete steps user choosing method that satisfies clean culture and multicast service simultaneously according to claim 1, described step 2) are as follows:

2-1) send unicast stream on the poorest user's the orthogonal subspaces at link and guarantee that unicast stream does not influence the poorest user performance of link by block diagonalization;

2-2) control unicast power makes the interference of clean culture can not make multicast produce after the stack clean culture than the worse user of the poorest user before superposeing;

Unicast power P _uCan not surpass P _{U, i}, P _{U, i}Satisfy

$\underset{i}{\min }\left\{\frac{P_{m,k}{\left|w_{m,k}^H{h}_i^{\left(k\right)}\right|}^2}{P_{u,i}\underset{j\&Element;U_k}{\mathrm{\&Sigma;}}{\left|w_{j,k}^H{h}_i^{\left(k\right)}\right|}^2+N_0}\right\}=\underset{i}{\min }\left\{\frac{P_m{\left|w_{m,k}^H{h}_i^{\left(k\right)}\right|}^2}{N_0}\right\}---\left(4\right)$

Unicast power can not surpass P _Max, the gross power of distributing to unicast service is P _u=min{P _{U, i}, P _Max, P wherein _MaxThe assignable maximum power of clean culture on the expression single carrier.

4. the user choosing method that satisfies clean culture and multicast service simultaneously according to claim 1, the concrete steps of described step 3) are as follows:

Travel through all possible unicast user collection, according to step 2) calculate the unicast rate under the various unicast user collection, select the unicast user collection of clean culture and speed maximum.

5. the user choosing method that satisfies clean culture and multicast service simultaneously according to claim 1 is promptly selected optimum unicast user collection, and the concrete steps of described step 3) are as follows:

According to step 2) at first select the maximum user of speed, selected to select next user on the basis this user then, if speed increase then continue to select, otherwise the user selects to finish.

Images