CN101442808B - United scheduling method for ascending multi-point collaboration in LTE-A - Google Patents

Abstract

The invention relates to an uplink multi-point coordinated joint scheduling method in LTE-A. The method includes the following steps: 1) determining cooperative and non-cooperative scheduling users; 2) intra-cell cooperative user scheduling; 3) inter-cell cooperative user joint resource scheduling; 4) intra-cell non-cooperative user scheduling. The present invention adopts the method of step-by-step optimization within the cell and between cells, and simultaneously considers the scheduling priority and channel correlation of users in the joint scheduling between cells, reduces the interference between co-frequency cooperative users, and ensures user fairness At the same time, it improves the throughput of the whole system and has a lower computational complexity.

Description

United scheduling method for ascending multi-point collaboration among a kind of LTE-A

Technical field

The invention belongs to moving communicating field, it relates to the united scheduling method for ascending multi-point collaboration among a kind of LTE-A.By adopting in the sub-district and the scheduling of minizone collaboration user federated resource, improve the Cell Edge User uplink throughput greatly.

Background technology

In existing LTE (Long Term Evolution, 3G Long Term Evolution), the throughput of Cell Edge User is well below central area user's throughput.In order to improve the data rate of Cell Edge User transmission, and the throughput and the average throughput efficiency that improve whole system, LTE-Advanced (Long TermEvolution-Advanced, senior Long Term Evolution, be abbreviated as LTE-A) with CoMP (Coordinated MultiPoint Transmission and Reception, multipoint cooperative) technology is classified one of its important candidate technologies as, and it mainly improves the cell edge throughput by the cooperation of minizone.

Multipoint cooperative can be divided between the base station multipoint cooperative in the multipoint cooperative and base station.Multipoint cooperative refers to user in each sub-district by a website (SITE) access network in the base station, can Data transmission or signaling between a plurality of websites, and manage by a base station or hub scheduler.What the present invention was primarily aimed at is the scheduling of the federated resource under the multipoint cooperative pattern in the base station, as shown in Figure 1, still can expand to the pattern of cooperation between base stations at an easy rate.

Common user's dispatching algorithm comprises polling algorithm and equitable proportion algorithm at present, and the equitable proportion method more has practical value in non-real-time service.In addition, NORTEL gives the method for carrying out multi-subscriber dispatching according to the subscriber channel spatial coherence.But these methods all realize in the sub-district.

In the CoMP system, the neighbor cell edge customer may occur with keeping pouring in defeated situation.If can not dispatch the user of neighbor cell well, will bring serious presence of intercell interference owing to the space correlation of channel.Therefore, how multiple sectors at same frequency user is carried out effective combined dispatching and become a problem demanding prompt solution.

Disturb the mode that suppresses to extract user data though can adopt, this also can bring very big capacitance loss.When there is very strong correlation in a plurality of collaboration users of frequency together between different districts among Fig. 1, for the information that the user is sent can correctly be received, just have to reduce the collaboration user number of same frequency sending, as shown in Figure 2, so greatly reduce throughput of system.

Summary of the invention

The present invention is for solving the above-mentioned technical problem that exists in the background technology, reduce greatly with disturbing between collaboration user frequently and provide a kind of, in guaranteeing the sub-district, improve throughput of system as far as possible in the equitable proportion between the user, and have the united scheduling method for ascending multi-point collaboration of the LTE-A system of lower complexity.The present invention relates to support the sub-district of cooperating, the sub-district of one group of support cooperation to become cooperation cell bunch; According to user's the channel quality or the business of order, the user of the edge and suitable Proportional Fair is called collaboration user, with central user or in real time the voice service user be called non-collaboration user; Collaboration user and non-collaboration user coexist in system.

Technical solution of the present invention is: the present invention is the united scheduling method for ascending multi-point collaboration among a kind of LTE-A, and its special character is: this method may further comprise the steps:

1) determines cooperation and non-cooperative scheduling user;

2) collaboration user scheduling in the sub-district;

3) minizone collaboration user federated resource scheduling;

4) non-collaboration user scheduling in the sub-district.

Above-mentioned steps 1) concrete steps are as follows:

1.1) determine according to the reference signal that receives whether user's signal quality arrives the requirement of access network; Determine to participate in user's set of scheduling;

1.2) according to the mensuration of reference signal strength and the demand of customer service, determine the type of dispatched users: cooperation or non-cooperation;

Above-mentioned steps 2) concrete steps are as follows:

2.1) according to the request rate and the average throughput of collaboration user, go out the order that each collaboration user is scheduled in the sub-district according to the equitable proportion algorithm computation, obtain the dispatching priority of collaboration user;

2.2) according to the dispatching priority and the channel information of collaboration user, be its allocated frequency band, according to priority order is assigned to collaboration user on the best frequency band of channel gain;

2.3) each sub-district in the cooperation cell bunch sends to hub scheduler or base station with dispatching priority and the frequency band division result who obtains respectively.

Above-mentioned steps 3) concrete steps are as follows:

3.1) if when the collaboration user of different districts all is in different frequency bands, direct execution in step 4 then); Otherwise execution in step 3.2);

3.2) calculate the channel relevancy between the collaboration user be on the same frequency band;

3.3) set the correlation threshold value, if channel relevancy value during, then direct execution in step 4) less than the correlation threshold value; Otherwise, adjust the frequency resource that collaboration user uses;

3.4) repeating step 3.1)～step 3.3), correlation is less than the correlation threshold value up between the different frequencies of all collaboration users or all same frequency collaboration users.

The concrete steps of the frequency resource that the adjustment collaboration user uses above-mentioned steps 3.3) are as follows:

3.3.1) if when the dispatching priority of each collaboration user in its Serving cell is identical, select the user of one or more sub-districts to carry out the frequency band fine setting arbitrarily, redistribute its service band according to the occupied situation of frequency band for the user;

3.3.2) if when user's dispatching priority is inequality, the low user of priority is carried out the frequency band fine setting.

Above-mentioned steps 3.2) the correlation calculations formula between two users is:

$R=\frac{1}{L}\underset{k=1}{\overset{L}{\mathrm{\Σ}}}\frac{\left|\right|H_{1k}{H}_{2k}^H\left|\right|}{\left|\right|H_{1k}\left|\right|\left|\right|H_{2k}\left|\right|}=\frac{1}{L}\underset{k=1}{\overset{L}{\mathrm{\Σ}}}{\left[\frac{\left|\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{h}_{i1}{h}_{i2}^{*}\right|}{{\left(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{\left|h_{i1}\right|}^2\right)}^{1/2}{\left(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{\left|h_{i2}\right|}^2\right)}^{1/2}}\right]}_k$

Wherein frequency bandwidth is L, and promptly the frequency number is L; H _1k=[h _I1] _kAnd H _2k=[h _I2] _k, represent two users respectively at the channel parameter of frequency k, h _IjExpression user j is to the channel parameter values between the reception antenna i.

Above-mentioned steps 4) concrete steps are as follows:

4.1) determine non-collaboration mode user's scheduling grade according to the Proportional Fair algorithm; And be the dispatching priority that each non-collaboration mode user determines frequency band;

4.2) be non-collaboration mode user allocated frequency band.

Above-mentioned steps 2.1) and step 4.1) in the equitable proportion algorithm as follows:

$\begin{array}{cc}\mathrm{Max}\frac{R_k\left(t\right)}{T_k\left(t\right)}& (k=\mathrm{1,2,3},...,M)\end{array}$

Wherein: R _k(t): the request rate of user k, T _k(t): user k is t in length of window _cThe average throughput of window

The more new formula of user's average throughput:

Calculate all users'

Be worth, and the user sorted according to the size of this value:

$\left\{\begin{array}{c}\frac{R_{P_1}\left(t\right)}{T_{P_1}\left(t\right)},\frac{R_{P_2}\left(t\right)}{T_{P_2}\left(t\right)},\frac{R_{P_3}\left(t\right)}{T_{P_3}\left(t\right)},\frac{R_{P_4}\left(t\right)}{T_{P_4}\left(t\right)}...\frac{R_{P_M}\left(t\right)}{T_{P_M}\left(t\right)}\\ b\\ P_1,P_2,P_3,P_4..{.P}_M\end{array}\right\}\⇒\left\{\begin{array}{c}\frac{R_1\left(t\right)}{T_1\left(t\right)},\frac{R_2\left(t\right)}{T_2\left(t\right)}...\frac{R_i\left(t\right)}{T_i\left(t\right)},\frac{R_j\left(t\right)}{T_j\left(t\right)}...\frac{R_M\left(t\right)}{T_M\left(t\right)}\\ b\\ Q_1,Q_2...Q_i,Q_j...Q_M\end{array}\right\}$

Wherein, { P ₁, P ₂, P ₃, P ₄... P _MBe unsorted dispatched users, { Q ₁, Q ₂... Q _i, Q _j... Q _MBe the dispatched users after the ordering, $\frac{R_i\left(t\right)}{T_i\left(t\right)}\&le;\frac{R_j\left(t\right)}{T_j\left(t\right)}(i<j).$

The present invention mainly is for the less a plurality of collaboration users of selective channel correlation between different districts send data simultaneously, when guaranteeing the fair transmission in each sub-district, improves the throughput of whole system.Its advantage is as follows:

1, in inter-cell united scheduling process,, greatly reduces between cooperation cell with the interference between the user frequently by considering cooperation cell with the channel relevancy between the frequency user.

2, the priority of user's scheduling in fair next initial each sub-district of proportion of utilization, and, in guaranteeing the sub-district, make user throughput high as far as possible in the equitable proportion between the user for the user selects best channel.

3, the stepped approach of the scheduling of non-collaboration user in the combined dispatching of the initial schedule by adopting collaboration user in each sub-district, minizone collaboration user and the sub-district is compared with all cooperation cell combined dispatching algorithms, has lower computation complexity.

Description of drawings

Fig. 1 is the interference between CoMP mode of operation and different user in the background technology;

Fig. 2 is the situation that CoMP deteriorates to can not have a plurality of users to send simultaneously on the identical frequency band;

Fig. 3 is a method flow diagram of the present invention;

Fig. 4 is definite cooperation and the non-cooperative scheduling user procedures flow chart among the present invention;

Fig. 5 is a collaboration user scheduling process flow chart in the sub-district among the present invention;

Fig. 6 is the minizone collaboration user combined dispatching process flow diagram among the present invention;

Fig. 7 is a non-collaboration user scheduling process flow chart in the sub-district among the present invention;

Bandwidth assignment schematic diagram in Fig. 8 sub-district after the scheduling of collaboration user initial resource;

Fig. 9 is the bandwidth assignment schematic diagram after the collaboration user federated resource scheduling of minizone;

Figure 10 is the bandwidth assignment schematic diagram after whole scheduling of resource process is finished.

Embodiment

Referring to Fig. 3, the present invention includes following steps:

1) user and the type thereof of selecting participation to dispatch according to channel status promptly determined cooperation and non-collaboration user;

2) collaboration user scheduling in the sub-district;

3) minizone collaboration user combined dispatching;

4) non-collaboration user scheduling in the sub-district.

Referring to Fig. 4, the method for definite cooperation of the present invention and non-collaboration user is as follows:

1.1) determine to participate in the dispatched users set, user's set can be expressed as: { K ₁, K ₂, K ₃, K ₄... K _M;

1.2) according to the mensuration of reference signal strength, judge user type: cooperation or non-cooperation; Wherein collaboration user can be expressed as: { C ₁, C ₂, C ₃, C ₄..., C _M1, non-collaboration user can be expressed as: { E ₁, E ₂, E ₃, E ₄... E _M2, constraint: M1+M2=M, wherein M1 is the collaboration user number, and M2 is non-collaboration user number, and M is the total number of users of usage ratio equity dispatching in the dispatching cycle;

Referring to Fig. 5, the collaboration user resource regulating method is as follows in the sub-district of the present invention:

2.1) according to user k rate request R _kWith its average throughput T _k, the proportion of utilization fair algorithm calculates the initial priority of user's scheduling;

The criterion of equitable proportion algorithm can be expressed as follows:

$\begin{array}{cc}\mathrm{Max}\frac{R_k\left(t\right)}{T_k\left(t\right)}& (k=\mathrm{1,2,3},...,M)\end{array}$

R _k(t): the request rate of user k

T _k(t): user k is t in length of window _cThe average throughput of window

The more new formula of user's average throughput:

Calculate all users'

Be worth, and the user sorted according to the size of this value:

$\left\{\begin{array}{c}\frac{R_{P_1}\left(t\right)}{T_{P_1}\left(t\right)},\frac{R_{P_2}\left(t\right)}{T_{P_2}\left(t\right)},\frac{R_{P_3}\left(t\right)}{T_{P_3}\left(t\right)},\frac{R_{P_4}\left(t\right)}{T_{P_4}\left(t\right)}...\frac{R_{P_M}\left(t\right)}{T_{P_M}\left(t\right)}\\ b\\ P_1,P_2,P_3,P_4..{.P}_M\end{array}\right\}\&DoubleRightArrow;\left\{\begin{array}{c}\frac{R_1\left(t\right)}{T_1\left(t\right)},\frac{R_2\left(t\right)}{T_2\left(t\right)}...\frac{R_i\left(t\right)}{T_i\left(t\right)},\frac{R_j\left(t\right)}{T_j\left(t\right)}...\frac{R_M\left(t\right)}{T_M\left(t\right)}\\ b\\ Q_1,Q_2...Q_i,Q_j...Q_M\end{array}\right\}$

Wherein, { P ₁, P ₂, P ₃, P ₄... P _MBe unsorted dispatched users, { Q ₁, Q ₂... Q _i, Q _j... Q _MBe the dispatched users after the ordering, $\frac{R_i\left(t\right)}{T_i\left(t\right)}\&le;\frac{R_j\left(t\right)}{T_j\left(t\right)}(i<j).$

2.2) be followed successively by the user according to user's dispatching priority and carry out frequency band division;

Select a user according to the order of dispatching priority, and select the unappropriated frequency band that channel quality is best for it;

2.3) user's schedule information is sent to hub scheduler or base station.

Single community user dispatching method when not considering to cooperate is compared, and the user in each sub-district dispatches main difference and is:

1, owing to consider the combination of multipoint cooperative, to judge at first that when scheduling of resource the user is that collaboration user also is non-collaboration user, be the collaboration user priority user of the selecting frequency range that shares out the work again;

2, each sub-district will report to the resource allocation conditions of collaboration user hub scheduler or base station, is used for the inter-cell united scheduling of back.

Referring to Fig. 6, minizone of the present invention collaboration user federated resource dispatching method is as follows:

3.1) if when the collaboration user of different districts all is in different frequency bands, direct execution in step 4 then); Otherwise execution in step 3.2).

3.2) calculate the adjacent sub-district of collaboration user of a sub-district with the channel relevancy between the frequency collaboration user;

For frequency bandwidth is the situation of L (being that the frequency number is L); H _1k=[h _I1] _kAnd H _2k=[h _I2] _kRepresent two users respectively at the channel parameter of frequency k, wherein h _IjRepresent user j to the channel parameter values between the reception antenna i,

Channel space correlation computing formula between two users is:

$R=\frac{1}{L}\underset{k=1}{\overset{L}{\mathrm{\&Sigma;}}}\frac{\left|\right|H_{1k}{H}_{2k}^H\left|\right|}{\left|\right|H_{1k}\left|\right|\left|\right|H_{2k}\left|\right|}=\frac{1}{L}\underset{k=1}{\overset{L}{\mathrm{\&Sigma;}}}{\left[\frac{\left|\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{h}_{i1}{h}_{i2}^{*}\right|}{{\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{\left|h_{i1}\right|}^2\right)}^{1/2}{\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{\left|h_{i2}\right|}^2\right)}^{1/2}}\right]}_k$

3.3) if channel relevancy value during less than threshold value, then direct execution in step 4);

Otherwise, adjust the frequency resource that collaboration user uses, the step of adjustment is as follows:

3.3.1) if when the dispatching priority of each collaboration user in its Serving cell is identical, select the user of one or more sub-districts to carry out the frequency band adjustment arbitrarily.Redistribute its service band according to the occupied situation of frequency band for the user.

3.3.2) if when user's dispatching priority is inequality, the low user of priority is carried out the frequency band fine setting.

3.4) repeating step 3.1)～step 3.3), correlation is less than threshold value up between the different frequencies of all collaboration users or all same frequency collaboration users.

Referring to Fig. 7, non-collaboration user resource regulating method is as follows in the sub-district of the present invention:

4.1) adopt and step 2.1) identical way is the non-collaboration user in each sub-district sort (dispatching priority);

4.2) be followed successively by non-collaboration user according to user's dispatching priority and carry out frequency band division;

Select a non-collaboration user according to the order of dispatching priority, and select the unappropriated frequency band that channel quality is best, up to finishing scheduling of resource for all users of this dispatching cycle for it

Referring to Fig. 8, the bandwidth assignment in the signal sub-district after the scheduling of collaboration user initial resource.

Referring to Fig. 9, the bandwidth assignment after the collaboration user federated resource scheduling of signal minizone.

Referring to Figure 10, the method for of the present invention 2 inter-cell united dispatching method specific embodiments is as follows:

1) 2 sub-districts (sub-district A and sub-district B) respectively have 4 collaboration users (A1, A2, A3, A4} and B1, B2, B3, B4}) and 4 non-collaboration users (with B5, B6, B7, B8});

2) the collaboration user frequency resource allocation is as shown in Figure 8 in the sub-district.Wherein collaboration user A1 and B1 are with frequently, and A3 and B2 are with frequently, and other collaboration users are frequently different;

3) since A1 and B1 in its sub-district separately all tool have plenty of the highest dispatching priority, therefore, we therefrom select one at random and carry out frequency adjustment, for example B1.Because the dispatching priority of A3 in its Serving cell is 3, and the dispatching priority of B2 in its Serving cell is 2, the dispatching priority of A3 is lower than the dispatching priority of B2, therefore B2 is adjusted.Through the frequency adjustment, B1 is adjusted in the cooperation cell scope not with user frequently, and A3 adjust to the very little B4 of its spatial coherence with frequently, therefore satisfied the requirement of inter-cell united scheduling, as shown in Figure 9.

4) sub-district A be A5, A6, A7, A8} dividing frequency resource, sub-district B is { B5, B6, B7, B8} dividing frequency resource.

The method according to this invention, existing method also can realize the federated resource scheduling of CoMP, and situation about comparing with inventive method is:

In the scheme of in the sub-district, dispatching, can pass through step 2) realize the frequency resource scheduling feature of multipoint cooperative, but when two collaboration user channel space correlations are very strong, can bring very big interference.

Have only in the method for a collaboration user in a plurality of sub-districts in a cooperative cluster, each collaboration user can not be with frequently, and this mode can reduce throughput of system owing to not utilizing space resources well.

Can also not adopt and dispatch and direct inter-cell united scheduling in the sub-district, this method can have optimum performance, but its complexity is can not be received in the actual realization by the reasonable optimizing algorithm.

Claims (4)

1. the united scheduling method for ascending multi-point collaboration among the LTE-A, it is characterized in that: this method may further comprise the steps:

1) determines according to the reference signal that receives whether user's signal quality arrives the requirement of access network; Determine to participate in user's set of scheduling; According to the mensuration of reference signal strength and the demand of customer service, determine cooperation or non-cooperative scheduling user;

2) according to the request rate and the average throughput of collaboration user, go out the order that each collaboration user is scheduled in the sub-district, obtain the dispatching priority of collaboration user according to the equitable proportion algorithm computation; According to the dispatching priority and the channel information of collaboration user, be its allocated frequency band, according to priority order is assigned to collaboration user on the best frequency band of channel gain; Each sub-district in the cooperation cell bunch sends to hub scheduler or base station with dispatching priority and the frequency band division result who obtains respectively;

3) minizone collaboration user federated resource scheduling, concrete steps comprise:

3.1) if when the collaboration user of different districts all is in different frequency bands, direct execution in step 4 then); Otherwise execution in step 3.2);

3.2) calculate the channel relevancy between the collaboration user be on the same frequency band;

3.3) set the correlation threshold value, if channel relevancy value during, then direct execution in step 4) less than the correlation threshold value; Otherwise, adjust the frequency resource that collaboration user uses;

3.4) repeating step 3.1)～step 3.3), correlation is less than the correlation threshold value up between the different frequencies of all collaboration users or all same frequency collaboration users;

4) respectively non-collaboration user is dispatched in each sub-district; Determine non-collaboration mode user's scheduling grade according to the Proportional Fair algorithm; And determine dispatching priority for each non-collaboration mode user, be non-collaboration mode user allocated frequency band.

2. the united scheduling method for ascending multi-point collaboration among a kind of LTE-A according to claim 1 is characterized in that: the concrete steps of the frequency resource that the adjustment collaboration user uses described step 3.3) are as follows:

3.3.1) if when the dispatching priority of each collaboration user in its Serving cell is identical, select the user of one or more sub-districts to carry out the frequency band fine setting arbitrarily, redistribute its service band according to the occupied situation of frequency band for the user;

3.3.2) if when user's dispatching priority is inequality, the low user of priority is carried out the frequency band fine setting.

3. the united scheduling method for ascending multi-point collaboration among a kind of LTE-A according to claim 2 is characterized in that: the correlation calculations formula described step 3.2) between two users is:

$R=\frac{1}{L}\underset{k=1}{\overset{L}{\mathrm{\&Sigma;}}}\frac{\left|\right|H_{1k}{H}_{2k}^H\left|\right|}{\left|\right|H_{1k}\left|\right|\left|\right|H_{2k}\left|\right|}=\frac{1}{L}\underset{k=1}{\overset{L}{\mathrm{\&Sigma;}}}{\left[\frac{\left|\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{h}_{i1}{h}_{i2}^{*}\right|}{{\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{\left|h_{i1}\right|}^2\right)}^{1/2}{\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{\left|h_{i2}\right|}^2\right)}^{1/2}}\right]}_k$

Wherein frequency bandwidth is L, and promptly the frequency number is L; H _1k=[h _I1] _kAnd H _2k=[h _I2] _k, represent two users respectively at the channel parameter of frequency k, h _IjExpression user j is to the channel parameter values between the reception antenna i.

4. the united scheduling method for ascending multi-point collaboration among a kind of LTE-A according to claim 1 is characterized in that: the equitable proportion algorithm is as follows described step 2) and in the step 4):

$\mathrm{Max}\frac{R_k\left(t\right)}{T_k\left(t\right)}$ K=1 wherein, 2,3 ..., M

Wherein: R _k(t): the request rate of user k, T _k(t): user k is t in length of window _cThe average throughput of window

The more new formula of user's average throughput:

Calculate all users' Be worth, and the user sorted according to the size of this value:

$\left\{\begin{array}{c}\frac{R_{P1}\left(t\right)}{T_{P1}\left(t\right)},\frac{R_{P2}\left(t\right)}{T_{P2}\left(t\right)},\frac{R_{P3}\left(t\right)}{T_{P3}\left(t\right)},\frac{R_{P4}\left(t\right)}{T_{P4}\left(t\right)}...\frac{R_{{\mathrm{PM}}_P}\left(t\right)}{T_{{\mathrm{PM}}_P}\left(t\right)}\\ b\\ P1,P2,P3,P4...{\mathrm{PM}}_P\\ \end{array}\right\}\&DoubleRightArrow;\left[\begin{array}{c}\frac{R_1\left(t\right)}{T_1\left(t\right)},\frac{R_2\left(t\right)}{T_2\left(t\right)}...\frac{R_i\left(t\right)}{T_i\left(t\right)},\frac{R_j\left(t\right)}{T_j\left(t\right)}...\frac{R_{M_P}\left(t\right)}{T_{M_P}\left(t\right)}\\ b\\ Q_1,Q_2...Q_i,Q_j...Q_{M_P}\\ \end{array}\right]$

Wherein, and P1, P2, P3, P4 ..., PM _PBe unsorted dispatched users,

Be the dispatched users after the ordering,

I＜j wherein.

Images