CN103281702B - A kind of collaboration communication method based on multiple cell dynamic clustering - Google Patents

Abstract

The invention discloses a kind of collaboration communication method based on multiple cell dynamic clustering, comprise the following steps: the first step: set up station list and calculate the matrix mutually disturbed minizone; Second step: the guiding base station selecting current cluster; Upgrade candidate list of base stations and selected station list; 3rd step: select one to the guiding base station i of current cluster _ldisturb maximum cell base station i ' to add current cluster, upgrade candidate list of base stations and selected station list; 4th step: return the 3rd step, until the number of base stations in current cluster equals set bunch size B _c; 5th step: carry out next bunch base station selected, the candidate list of base stations after once upgrading as of late is empty set 6th step: all bunches that in system, all base stations are set up by the 5th step select station list, carry out collaboration communication.This communication means makes each base station determine whether and its cluster according to the degree of association size with other base stations, greatly accelerates dynamic clustering speed, thus improves communication efficiency.<pb pnum="1" />

Description

A kind of collaboration communication method based on multiple cell dynamic clustering

Technical field

The invention belongs to the wireless cellular network communications field, specifically, relate to a kind of collaboration communication method based on multiple cell dynamic clustering.

Background technology

Along with the sharp increase of userbase and increasing of wireless traffic, cellular communications networks faces a significant challenge---improve the utilance of limited band resource.For satisfying the demands, base station will be set up more thick and fast, frequency repeat utilization ratio will improve, and even takies the available band of whole distribution.But this will cause more serious presence of intercell interference (inter-cell interference, ICI) further, make QoS of customer (Quality ofService, QoS) be affected, particularly Cell Edge User.Collaboration communication is a kind of effective means being devoted to alleviate presence of intercell interference, and 3GPP is formal includes cooperative communication technology in LTE-Advanced standard, and making it becomes the key technology improving future wireless network systematic function.

Collaboration communication is that abundant development system spatial degrees of freedom provides possibility, and its basic thought is in the downlink, and multiple community to the combined signal process that will send, thus makes home cell suppress the interference signal of other communities.Base station cooperation in a cooperation unit forms one bunch, and like this by cooperation, in bunch, the interference of minizone will be effectively suppressed.In theory, more with number of base stations in cluster, presence of intercell interference is suppressed more obvious; But in practicality, more cell cooperative means more feedback overhead, computing cost and backbone network transport overhead, this situation makes number of cells in same cluster can not be too many.

For choosing which this problem of base station cluster, some fruitful algorithms are suggested, mostly it is considered that the static policies of foundation base station geographic close position cluster, namely no longer change behind selected cluster base station, but wireless channel time-varying characteristics may make this method improve limited to performance.To this, the base station of some dynamic clustering methods---cluster constantly changes according to channel information and is suggested, and systematic function relative quiescent clustering method is had clear improvement.But, existing dynamic clustering algorithm be mostly all clusters may in search thoroughly and calculate have maximum system capacity bunch as a result, this process is quite complicated; There is scholar to make slight change to this kind of algorithm, calculate target and be changed to the channel gain of all possible bunch and the cluster result selecting overall gain maximum; But these methods all need the possibility of all clusters of limit.When network size is larger, even if base station number is less in every cluster, amount of calculation is also quite large, and after may having calculated, channel condition changes already.Therefore, it is inapplicable when this exhaustive search method is in real network.

Summary of the invention

Technical problem: technical problem to be solved by this invention is: a kind of collaboration communication method based on multiple cell dynamic clustering is provided, this communication means makes each base station can determine whether and its cluster according to the degree of association size with other base stations, greatly accelerate dynamic clustering speed, thus improve communication efficiency.

Technical scheme: for solving the problems of the technologies described above, the technical solution used in the present invention is:

Based on a collaboration communication method for multiple cell dynamic clustering, this multiple cell comprises at least three communities, is provided with travelling carriage and base station in each community, and this communication means comprises the following steps:

The first step: set up station list and calculate the matrix mutually disturbed minizone:

Initialization candidate list of base stations Candidacy={1,2 ..., B}, selects station list b represents the sum of base station in multiple cell, sets the base station maximum quantity comprised in each bunch and is B _c, B _c< B, the number in multiple cell bunch is wherein, symbol represent and round downwards;

The channel information matrix H of each base station is fed back to, by formula (1) the measuring and calculating mutual interference matrix F in minizone according to all travelling carriages of multiple cell:

$f_{\mathrm{ij}}=\left\{\begin{array}{c}{\left|h_{\mathrm{ij}}\right|}^2+{\left|h_{\mathrm{ji}}\right|}^2,i\&NotEqual;j\\ 0,i=j\end{array}\right.$ Formula (1)

Wherein, h _ijrepresent the channel gain between user i and the transmitting antenna of base station j, h _jirepresent the channel gain between user j and the transmitting antenna of base station i, f _ijfor the i-th row jth column element of mutual interference matrix F, f _ijrepresent the mutual interference degree between base station i and base station j;

Second step: select a cell base station i being subject to other area interference maximum in candidate list of base stations _l, as the guiding base station of current cluster, guide base station i _lmeet formula (2):

$i_l=\underset{i\&Element;\mathrm{Candidacy}}{\max }\underset{j\&NotEqual;i}{\mathrm{\&Sigma;}}{f}_{\mathrm{ij}}$ Formula (2)

According to formula (3) by i _ldelete from candidate list of base stations, thus upgrade candidate list of base stations, according to formula (4) by i _ladds to and select in station list, thus renewal selects station list:

Candidacy1=Candidacy-{i _lformula (3)

Preclusion1=Preclusion+{i _lformula (4)

Wherein, Candidacy1 represent first upgrade after candidate list of base stations, Preclusion1 represent first upgrade after select station list;

3rd step: select a guiding base station i to current cluster in candidate list of base stations in the updated _ldisturb maximum cell base station i ' to add current cluster, base station i ' meets f _{ii '}represent the annoyance level between base station i and base station i '; Base station i ' is deleted from the candidate list of base stations after last update, upgrades candidate list of base stations; Base station i ' is added to and selects in station list after last update, upgrade and select station list, then enter the 4th step;

4th step: return the 3rd step, until the number of base stations in current cluster equals B _c, then enter the 5th step;

5th step: by the candidate list of base stations alternatively station list after the 4th step upgrades, using through the 4th step upgrade after select station list as selecting station list, return second step, carry out next bunch base station selected, the candidate list of base stations after once upgrading as of late is empty set after last update select station list be 1,2 ..., B}, to set up in multiple cell all bunches select station list;

6th step: all bunches that in system, all base stations are set up by the 5th step select station list, carry out collaboration communication by cluster result.

Beneficial effect: compared with prior art, technical scheme of the present invention has following beneficial effect:

1. calculate interaction matrix simple to operate, only a sub-addition just can obtain, and this matrix accurately can weigh the degree that influences each other between community, and non-individual is on the impact of the other side.

2. adopt and guide base station selecting method, in bunch, different base station has different priority, eliminate the randomness of conventional dynamic clustering method to a certain extent, improve the information rate performance of collaboration communication, the formation accurately selecting influenced maximum cell base station to dominate bunch.

3., to configure different priority to guiding base station cell annoyance level for other base stations, finally eliminate the randomness of conventional dynamic clustering method, thus make cluster bunch in disturb minimum.The present invention determines whether and its cluster according to the degree of association size of base station and other base stations, greatly accelerates dynamic clustering speed.

Accompanying drawing explanation

Fig. 1 is the system schematic of the embodiment of the present invention, and wherein, solid line represents useful signal link, and dotted line represents interference signal link, the base station cluster in light-colored part Liang Ge community, the base station cluster in dark parts three communities.

Fig. 2 be bunch in base station number B _cwhen=2, the method for the embodiment of the present invention, static clustering method and the contrast string diagram of system average size obtained respectively without cooperation clustering method.

Fig. 3 be bunch in base station number B _cwhen=3, the method for the embodiment of the present invention, static clustering method and the contrast string diagram of system average size obtained respectively without cooperation clustering method.

Fig. 4 be bunch in base station number B _cwhen=4, the method for the embodiment of the present invention, static clustering method and the contrast string diagram of system average size obtained respectively without cooperation clustering method.

Fig. 5 be bunch in base station number B _cwhen=2, the method for the embodiment of the present invention, tradition are based on the dynamic clustering method of searching mode thoroughly and the contrast string diagram of system average size obtained respectively without cooperation clustering method.

Fig. 6 be bunch in base station number B _cwhen=3, the method for the embodiment of the present invention, tradition are based on the dynamic clustering method of searching mode thoroughly and the contrast string diagram of system average size obtained respectively without cooperation clustering method.

Fig. 7 be bunch in base station number B _cwhen=4, the method for the embodiment of the present invention, tradition are based on the dynamic clustering method of searching mode thoroughly and the contrast string diagram of system average size obtained respectively without cooperation clustering method.

Embodiment

Below in conjunction with drawings and Examples, technical scheme of the present invention is described in further detail.

The multiple cell that the present invention relates to comprises at least three communities, is provided with travelling carriage and base station in each community.A kind of collaboration communication method based on multiple cell dynamic clustering of the present invention, comprises the following steps:

The first step: set up station list and calculate the matrix mutually disturbed minizone:

Initialization candidate list of base stations Candidacy={1,2 ..., B}, selects station list b represents the sum of base station in multiple cell, sets the base station maximum quantity comprised in each bunch and is B _c, B _c< B, the number in multiple cell bunch is wherein, symbol represent and round downwards;

The channel information matrix H of each base station is fed back to, by formula (1) the measuring and calculating mutual interference matrix F in minizone according to all travelling carriages of multiple cell:

$f_{\mathrm{ij}}=\left\{\begin{array}{c}{\left|h_{\mathrm{ij}}\right|}^2+{\left|h_{\mathrm{ji}}\right|}^2,i\&NotEqual;j\\ 0,i=j\end{array}\right.$ Formula (1)

Wherein, h _ijrepresent the channel gain between user i and the transmitting antenna of base station j, h _jirepresent the channel gain between user j and the transmitting antenna of base station i, f _ijfor the i-th row jth column element of mutual interference matrix F, f _ijrepresent the mutual interference degree between base station i and base station j.Preferably, B _c{ value in 2,3,4}, the quantity of B is less than or equal to 37.

Second step: select a cell base station i being subject to other area interference maximum in candidate list of base stations _l, as the guiding base station of current cluster, guide base station i _lmeet formula (2):

$i_l=\underset{i\&Element;\mathrm{Candidacy}}{\max }\underset{j\&NotEqual;i}{\mathrm{\&Sigma;}}{f}_{\mathrm{ij}}$ Formula (2)

According to formula (3) by i _ldelete from candidate list of base stations, thus upgrade candidate list of base stations, according to formula (4) by i _ladds to and select in station list, thus renewal selects station list:

Candidacy1=Candidacy-{i _lformula (3)

Preclusion1=Preclusion+{i _lformula (4)

Wherein, Candidacy1 represent first upgrade after candidate list of base stations, Preclusion1 represent first upgrade after select station list;

3rd step: select a guiding base station i to current cluster in candidate list of base stations in the updated _ldisturb maximum cell base station i ' to add current cluster, base station i ' meets f _{ii '}represent the annoyance level between base station i and base station i '.Base station i ' is deleted from the candidate list of base stations after last update, upgrades candidate list of base stations; Base station i ' is added to and selects in station list after last update, upgrade and select station list, then enter the 4th step.

4th step: return the 3rd step, until the number of base stations in current cluster equals B _c, then enter the 5th step.

5th step: by the candidate list of base stations alternatively station list after the 4th step upgrades, using through the 4th step upgrade after select station list as selecting station list, return second step, carry out next bunch base station selected, the candidate list of base stations after once upgrading as of late is empty set after last update select station list be 1,2 ..., B}, to set up in multiple cell all bunches select station list;

6th step: all bunches that in system, all base stations are set up by the 5th step select station list, carry out collaboration communication by cluster result.

Collaboration communication method based on multiple cell dynamic clustering provided by the invention, for reducing the common-channel interference of minizone, LTE-A standard introduces a kind of mechanism of multi-cell cooperating, by multiple base stations associated treatment receiving and transmitting signal the set of the base station in an IWU (be called bunch), thus to become interference signal be useful signal.For how to choose this problem of base station cluster, existing method generally according to geographical far and near as sole criterion between base station, cannot adapt to wireless channel time-varying characteristics.Dynamic clustering method allows different base station cluster adaptively according to channel variation, existing dynamic clustering method all by search thoroughly solve make channel gain maximum cluster may, the huge amount of calculation of this mode hinders that it is practical.The friend-making mechanism of communication means simulating human of the present invention, proposes the dynamic clustering method based on friend-making algorithm.Relatively existing dynamic clustering method, dynamic clustering method of the present invention has the low advantage of complexity.Method of the present invention constructs the mutual interference matrix of each minizone according to the channel condition information that each travelling carriage feeds back, and which base station cluster the annoyance level between each community of direct basis determines, disposablely can try to achieve cluster result.The base station collaboration cluster that this calculation side makes common-channel interference serious, offsets interference by ZF precoding, thus improves power system capacity.The present invention is based on the characteristic of mankind's make friend activity, be applied to multi-cell cooperating dynamic clustering problem, each base station can be determined whether and its cluster according to the degree of association size with other base stations, greatly accelerate cluster speed, and gained performance is almost with to search mode thoroughly identical, and this method will be well suited for practical application.

Community as shown in Figure 1, adopts the collaboration communication pattern of method establishment of the present invention.Five communities are provided with in Fig. 1, solid line represents useful signal link, dotted line represents interference signal link, the base station of light-colored part Liang Ge community (being namely positioned at leftmost Liang Ge community) becomes one bunch, one, the base station cluster of dark parts three communities (being namely positioned at three communities on middle part and the right).

Exemplify an embodiment below.

This embodiment chooses four layers of cell pattern, and in system, total number of base is 37, i.e. B=37, number B in base station in bunch _cin { value in 2,3,4}.Arranging radius of society is 1km, and travelling carriage is randomly dispersed in each position in community, and base station adopts polling schemas schedule mobile stations, and in bunch, base station number is fixed as B _c, remain less than B _cthe independent cluster in individual base station.Channel model between travelling carriage i and base station j is shown below:

$h_{\mathrm{ij}}={\mathrm{\&Gamma;}}_{\mathrm{ij}}\sqrt{G_{\mathrm{ij}}\mathrm{\&beta;}{d}_{\mathrm{ij}}^{-\mathrm{\&alpha;}}{r}_{\mathrm{ij}}}$

Wherein, Γ _ijbe variance be 1, obey the Rayleigh fading of multiple Gaussian Profile, G _ijfor dual-mode antenna gain is amassed, α is path loss index, and β is path loss constant, d _ijthe distance between travelling carriage i and base station j, r _ijfor obeying the shadow fading of logarithm Gaussian Profile.According to existing 3GPP give model (source is seen: LTE-AEvaluation Model, NTT DocoMo, Available online.), G is set _ij=9dB, α=3.76, β=10 ^-14.84, r _ijaverage be 0dB, variance is 8dB.Given antenna for base station maximum transmission power is P _n=1W, definition relative signal-to-noise ratio is the signal to noise ratio that cell edge only considers path loss and Gaussian noise.ZF precoding is adopted to carry out 3000 Monte-Carlo Simulation.Simulation result is as shown in table 1, Fig. 2 to Fig. 7.Clustering method as a comparison has: (source is seen: S.Zhou to search tactful dynamic clustering method thoroughly, J.Gong, Z.Niu, Y.Jia, and P.Yang, " A decentralized framework fordynamic downlink base station cooperation, " in Proc.IEEE GLOBECOM, Dec.2009, pp.1 – 6.), static clustering method and the clustering method (source see Wang L C.3-CellNetwork MIMO Architectures with Sectorization and Fractional FrequencyReuse [J] .IEEE Journal on Selected Areas in Communications.2011.29 (6): 1185-1199.) without cooperation.

Table 1

In Table 1, T _mrepresent the CM number of times in communication means of the present invention needed for cluster, T _arepresent that the plural number in communication means of the present invention needed for cluster adds number of times, T _crepresent the number of times of the comparison operation in communication means of the present invention needed for cluster; T _mthe CM number of times needed for dynamic clustering method of the poor search method mentioned in ' expression background technology, T _a' represent that the plural number needed for dynamic clustering method of the poor search method mentioned in background technology adds number of times, T _cthe number of times of the comparison operation needed for dynamic clustering method of the poor search method mentioned in ' expression background technology.As can be seen from Table 1, when having number of base stations in same cluster, T _mmuch smaller than T _m', T _amuch smaller than T _a', T _cmuch smaller than T _c'.

As can be seen from Fig. 2 to Fig. 4, compared with static clustering method, adopt dynamic clustering method of the present invention, the average size gain that system obtains obtains larger must raising.As can be seen from Fig. 5-Fig. 7, with based on search thoroughly strategy dynamic clustering method compared with, adopt the current performance of dynamic clustering method of the present invention to be more or less the same, but its amount of calculation obviously reduces.

Claims (2)

1. based on a collaboration communication method for multiple cell dynamic clustering, it is characterized in that, this multiple cell comprises at least three communities, is provided with travelling carriage and base station in each community, and this communication means comprises the following steps:

The first step: set up station list and calculate the matrix mutually disturbed minizone:

Initialization candidate list of base stations Candidacy={1,2 ..., B}, selects station list b represents the sum of base station in multiple cell, sets the base station maximum quantity comprised in each bunch and is B _c, B _c< B, the number in multiple cell bunch is wherein, symbol expression rounds up;

The channel information matrix H of each base station is fed back to, by formula (1) the measuring and calculating mutual interference matrix F in minizone according to all travelling carriages of multiple cell:

$f_{\mathrm{ij}}=\left\{\begin{array}{cc}{\left|h_{\mathrm{ij}}\right|}^2+{\left|h_{\mathrm{ji}}\right|}^2,& i\&NotEqual;j\\ 0,& i=j\end{array}\right.$ Formula (1)

Wherein, h _ijrepresent the channel gain between user i and the transmitting antenna of base station j, h _jirepresent the channel gain between user j and the transmitting antenna of base station i, f _ijfor the i-th row jth column element of mutual interference matrix F, f _ijrepresent the mutual interference degree between base station i and base station j;

Second step: select a cell base station i being subject to other area interference maximum in candidate list of base stations _l, as the guiding base station of current cluster, guide base station i _lmeet formula (2):

$i_l=\underset{i\&Element;\mathrm{Candidacy}}{\max }\underset{j\&NotEqual;i}{\mathrm{\&Sigma;}}{f}_{\mathrm{ij}}$ Formula (2)

According to formula (3) by i _ldelete from candidate list of base stations, thus upgrade candidate list of base stations, according to formula (4) by i _ladds to and select in station list, thus renewal selects station list:

Candidacy1=Candidacy-{i _lformula (3)

Preclusion1=Preclusion+{i _lformula (4)

Wherein, Candidacy1 represent first upgrade after candidate list of base stations, Preclusion1 represent first upgrade after select station list;

3rd step: select a guiding base station i to current cluster in candidate list of base stations in the updated _ldisturb maximum cell base station i ' to add current cluster, base station i ' meets f _{ii '}represent the annoyance level between base station i and base station i '; Base station i ' is deleted from the candidate list of base stations after last update, upgrades candidate list of base stations; Base station i ' is added to and selects in station list after last update, upgrade and select station list, then enter the 4th step;

4th step: return the 3rd step, until the number of base stations in current cluster equals B _c, then enter the 5th step;

5th step: by the candidate list of base stations alternatively station list after the 4th step upgrades, using through the 4th step upgrade after select station list as selecting station list, return second step, carry out next bunch base station selected, the candidate list of base stations after once upgrading as of late is empty set after last update select station list be 1,2 ..., B}, to set up in multiple cell all bunches select station list;

6th step: all bunches that in system, all base stations are set up by the 5th step select station list, carry out collaboration communication by cluster result.

2. according to the collaboration communication method based on multiple cell dynamic clustering according to claim 1, it is characterized in that, described B _c{ value in 2,3,4}, the quantity of B is less than or equal to 37.