CN102118754B - Partitioning method of dynamic cooperation sets in CoMP technology - Google Patents

Abstract

The invention discloses a partitioning method of dynamic cooperation sets in CoMP technology, comprising the following steps: firstly, establishing a system model which is arranged in a cellular mobile telecommunication network and is provided with B base stations and 3 multiplied by B cells, in which NU users are distributed uniformly; secondly, determining cooperative users according to a threshold method, setting a signal-to-interference-noise threshold (SINRthreshold), and sorting the users into the cooperative users if the signal-to-interference-noise ratio of the user is less than the SINRthreshold; thirdly, determining the cooperation sets; and finally, determining the cooperative users. In the partitioning method, cell-edge users can be positioned in the reasonable cooperation sets, the same-frequency interference in cells can be effectively eliminated, and the throughput of the cell-edge users is improved, so that the performance of a whole system is improved.

Description

The division methods of dynamic cooperative collection in a kind of CoMP

Technical field

The invention belongs to wireless communication technology field, the division methods of dynamic cooperative collection in especially a kind of CoMP technology.The method can be the in the situation that of little feedback overhead and low complex degree, and the collaboration set of reasonably optimizing Cell Edge User reduces the suffered interference of Cell Edge User, improves the throughput of Cell Edge User, thereby improves the performance of system.

Background technology

In traditional mobile communication, the user in this community is only responsible in each community, not cooperation between community.There is co-channel interference in the intersection of the situation Xia， community that is 1 at frequency duplex factor as one, causes the communication quality of Cell Edge User to decline, thereby the spectrum efficiency of whole communication system is reduced.

In the meeting of the wireless access network of holding at 3GPP (TSG-RAN) working group 1 (WGl), in order to solve above-mentioned run into problem, CoMP technology is suggested as a key technology of LTE_Advanced.CoMP technology is coordinate multipoint transmitting/receiving technology, and the multiple spot here refers to geographically separated a plurality of antenna access points, can be base station, user or via node.Between base station, with optical fiber, connect (X2 interface), by X2 interface, share control information and business information; Several antennas station in same collaboration set or node provide data, services for a Cell Edge User simultaneously, thereby can effectively solve the problem of co-channel interference of Cell Edge User, arrive the object that improves systematic function.

Compositing characteristic from CoMP transmission technology, in order to reduce minizone co-channel interference to reach better systematic function, need to study for the aspects such as division, wireless resource scheduling algorithm and MIMO definite, collaboration set of the cooperative node selection in CoMP transmission technology, collaboration user.Division methods mainly for the collaboration set in CoMP transmission technology has proposed a kind of new algorithm herein, and the user in cooperation state is selected to a rational cooperation transmission set, and the transmission performance of maximized raising system, reduces the interference of minizone.

Aspect collaboration set division, mainly containing two kinds of methods: the collaboration set based on UE is divided and the collaboration set based on Cell is divided.In the collaboration set based on UE is divided, UE is according to the collaboration set of the Location of requirement of self oneself, profit can be brought larger performance boost for system in this way, but in the case, in system, the collaboration set of different UEs can be overlapping, making the complexity of scheduling of resource very high, is infeasible in actual applications.In the collaboration set based on Cell is divided, collaboration set be take community as base unit, and collaboration set Nei community provides service for the edge customer in collaboration set simultaneously, by rational collaboration set, divides, and can effectively improve the overall performance of system.This method is for the collaboration set based on UE is divided, overlapped due to what cooperation was not gathered, make realizing of source scheduling fairly simple, and it is suitable in the collaboration set aspect the performance boost of system with based on UE, dividing, so the division of the collaboration set of this paper has adopted the collaboration set division methods based on Cell.

Aspect collaboration set division, conventional method is that static collaboration set is divided at present.The method is divided into a fixing collaboration set by contiguous microzonation in the process communicating, as shown in Figure 1, this method is not optimized collaboration set, and the residing collaboration set of each Cell Edge User is also not exclusively reasonable, can not bring into play the advantage of cooperation completely.Therefore in CoMP technology, be necessary collaboration set reasonably to optimize, just can make each user in rational collaboration set.Dynamic cooperative collection division methods in this paper, from whole optimizing cells angle, in the situation that little feedback overhead and low complex degree are selected rational collaboration set according to the demand of all collaboration users in community, the residing collaboration set of Cell Edge User is optimized.

Summary of the invention

The object of the invention is in order to overcome the irrationality that in existing CoMP technology, collaboration set is divided, the in the situation that of little feedback overhead and low complex degree, a kind of dynamic cooperative collection division methods has been proposed, this division methods can make Cell Edge User in rational collaboration set, can effectively eliminate the co-channel interference of minizone, improve the throughput of Cell Edge User, thereby improve the performance of whole system.

The object of the invention is to solve by the following technical programs:

The division methods of dynamic cooperative collection in this kind of CoMP technology, specifically comprises the following steps:

1) set up system model

Be located in cellular mobile communication networks, have B base station, 3 * B community, each community is uniform-distribution with N _uindividual user;

2) by threshold method, determine collaboration user

Set Signal to Interference plus Noise Ratio thresholding SINR _thresholdif user's Signal to Interference plus Noise Ratio is less than SINR _threshold, put user under collaboration user;

3) determine collaboration set

(3.1) collaboration set of establishing in whole cellular mobile communication networks forms set Ф, and now Ф is empty set;

(3.2) in cellular mobile communication networks, all communities that have neither part nor lot in cooperation form set omega, and in set omega, all possible collaboration set forms set

wherein

represent k standby collaboration set in Ψ, here c _{i, k}for standby collaboration set

in i community, N _crepresent the number of cells in each collaboration set, then utilize formula (1) to calculate the weights of all standby collaboration sets, collaboration set

weights V _kbe expressed as:

In formula, I _krepresent collaboration set

interior Cell Edge User sum,

represent Gaussian noise power, represent collaboration set

interior cell i is to the large scale decline of Cell Edge User k;

(3.3) from set

middle selection collaboration set weights V _kmaximum collaboration set

if set Ф all collaboration set Zhong community all with collaboration set

zhong community is different, just by standby collaboration set put under in set Ф;

(3.4) finally order set

return step (3.2), operation is as above carried out in all communities in Ω, until the number of cells in Ω is zero; By the division of dynamic cooperative collection, be divided into altogether M collaboration set, M collaboration set forms set Ф={ φ ₁, φ ₂..., φ _m..., φ _m, φ wherein _mbe m collaboration set.

channel condition information by long-time statistical terminal feedback to base station obtains, and according to

variation collaboration set is revised dynamically, to meet the demand of Cell Edge User now.

4) finally determine collaboration user

Calculate the Signal to Interference plus Noise Ratio SINR of each collaboration user, select new Signal to Interference plus Noise Ratio thresholding, the user that SINR is lower than new Signal to Interference plus Noise Ratio thresholding puts non-collaboration user under, finally collaboration user number is controlled to 10% to 20%.

Above-mentioned steps 3) in, user under collaboration situation, collaboration set φ _mthe reception signal indication of interior k collaboration user is:

$y_{k,m}=\underset{m\∈{\mathrm{\φ}}_k}{\mathrm{\Σ}}\sqrt{p_m}{H}_{k,m}{W}_{k,m}{s}_{k,m}+\underset{n\∉{\mathrm{\φ}}_k}{\mathrm{\Σ}}\underset{w=1}{\overset{N_U}{\mathrm{\Σ}}}\sqrt{p_n}{H}_{k,n}{W}_{w,n}{s}_{w,n}+B---\left(2\right)$

In formula, y _{k, m}represent collaboration set φ _mthe reception signal of interior k Cell Edge User, p _mfor the transmitting power of community m to user k, s _{k, m}for the transmitted signal of community m to user k, H _{k, m}that community m is to the channel matrix of user k, W _{k, m}be community m to the pre-coding matrix of user k, B is Gaussian noise, N _ufor the number of each community user, in formula, first is Serving cell and the cooperation cell useful signal to user k transmission, and second is the interference to this user of other community outside collaboration set.

The present invention has following beneficial effect:

The present invention is first by the analysis to channel component characteristic and feedback overhead, the large scale that obtains channel by long-time statistical declines, the feedback overhead of this method is few, the feedback cycle of general channel condition information is 200～1000 TTI, do not need in real time, change fast, and the channel large scale that long-time statistical is obtained decline is applied to during collaboration set divides; Then from the angle of whole system optimization, with the decline of channel large scale, calculate the weights of all collaboration users in collaboration set; The weights that last basis obtains are determined most suitable cooperation cell in collaboration set, and the collaboration set of Cell Edge User is optimized.This method, the in the situation that of little feedback overhead and low complex degree, can reduce the suffered interference of Cell Edge User, improves the spectrum efficiency of Cell Edge User, thereby improves the performance of system.

Accompanying drawing explanation

Fig. 1 is dynamic cooperative collection partiting step schematic diagram;

Fig. 2 is that dynamic cooperative collection is divided flow chart;

Fig. 3 is a kind of dynamic cooperative collection division result.

Embodiment

Below in conjunction with accompanying drawing, the present invention is described in further detail with concrete embodiment.

In cellular mobile communication networks, if the edge of user in community, the interference that is subject to adjacent cell is larger.In order to eliminate the interference of adjacent cell, need adjacent cell and Serving cell to form a collaboration set, be Cell Edge User service jointly.The present invention is the method that the residing collaboration set of Cell Edge User is dynamically divided and optimized.

While carrying out the division of dynamic cooperative collection, as shown in Figure 1, comprise the steps:

1) set up system model

Be located in cellular mobile communication networks, have B base station, 3 * B community, each community is uniform-distribution with N _uindividual user.Total M collaboration set in whole cellular system, all collaboration sets form gathers Ф, suc as formula shown in: Ф={ φ ₁, φ ₂..., φ _m... φ _m, in formula, φ _mrepresent m collaboration set, φ _min have N _cindividual community and I _mindividual Cell Edge User, total UE in system _coMP=I ₁+ I ₂+ ... + I _mindividual Cell Edge User.

The signal indication that user k receives in non-collaboration situation Xia， community c is:

$y_{k,c}=\sqrt{p_c}{H}_{k,c}{W}_{k,c}{s}_{k,c}+\underset{n=1,n\≠c}{\overset{3\×B}{\mathrm{\Σ}}}\underset{w=1}{\overset{N_U}{\mathrm{\Σ}}}\sqrt{p_n}{H}_{k,n}{W}_{w,n}{s}_{w,n}+n$

In formula, y _{k, c}the reception signal that represents user k in the c of community, s _{k, c}for the transmitted signal of community c to user k, H _{k, c}and W _{k, c}that community c is to channel matrix and the pre-coding matrix of user k, P _cfor the transmitting power of community c to user k, n is Gaussian noise.In formula, first is useful signal, second interference signal that is minizone.

Under collaboration situation, collaboration set φ _mthe reception signal indication of an interior k collaboration user is:

$y_{k,m}=\underset{c\∈{\mathrm{\φ}}_m}{\mathrm{\Σ}}\sqrt{p_c}{H}_{k,c}{W}_{k,c}{s}_{k,c}+\underset{n\∉{\mathrm{\φ}}_m}{\mathrm{\Σ}}\underset{w=1}{\overset{N_U}{\mathrm{\Σ}}}\sqrt{p_n}{H}_{k,n}{W}_{w,n}{s}_{w,n}+n$

In formula, y _{k, m}represent collaboration set φ _mthe reception signal of interior k collaboration user.First is Serving cell and the cooperation cell useful signal to user k transmission, and second is the interference to this user of other community outside collaboration set.

2) by threshold method, determine collaboration user

Set letter dry than thresholding SINR _thresholdif user's letter is dried and is less than SINR than SINR _threshold, user being put under to collaboration user (is SINR < SINR _threshold); The threshold value here should be slightly larger according to the characteristic setting of system, allows so the lower user of SINR as much as possible participate in determining of collaboration set.

The Signal to Interference plus Noise Ratio SINR of user k in cell i now _kcomputing formula as follows:

${\mathrm{SINR}}_k=\frac{{\left|{\hat{h}}_{k,i}\right|}^2}{\underset{j\&NotEqual;i}{\mathrm{\&Sigma;}}{\left|{\hat{h}}_{k,j}\right|}^2+{\mathrm{\&sigma;}}_n^2}$

In formula,

represent that cell i is to the channel large scale decline of user k, represent Gaussian noise power;

3) definite dynamic cooperative collection

Collaboration set φ _min all Cell Edge User the separately the channel condition information on the subcarrier of place be:

In formula, H _{i, j, n}be illustrated on n subcarrier community j to the multipath fading matrix of Cell Edge User i, Γ (d _{i, j}) represent that community j is to the large scale decline (comprising path loss and shadow fading) of Cell Edge User i, d _{i, j}represent that the base station of cell i is to the distance of Cell Edge User i.

In formula (a), the squared envelope of compound channel meets gamma logarithm normal distribution, can be approximately logarithm normal distribution.The effect of multipath fading is to lower the average of Compound Distribution and increase variance, this approximate good effect of playing when the standard deviation of shadow fading or the shaping factor increase of Nakagami distribution.

According to the geographical distribution of antenna for base station, the multipath fading between different base station is incoherent, and the spatial coherence between the antenna in base station is very little on the impact of whole spatial coherence.In formula (a), less to signal amplitude influence of fading due to multipath fading, utilize the slow property of large scale decline can effectively reduce feedback overhead simultaneously, therefore utilize the large scale decline of channel to determine collaboration set herein, formula in the situation that not considering subcarrier difference (a) can be simplified to:

In formula,

represent collaboration set φ _minterior community j is to the large scale decline of Cell Edge User k.

From formula (b), the SINR of Cell Edge User k _kfor:

${\mathrm{SINR}}_k=\frac{\underset{i\&Element;{\mathrm{\&phi;}}_m}{\mathrm{\&Sigma;}}{\left|{\hat{h}}_{k,i}\right|}^2}{\underset{j\&NotEqual;{\mathrm{\&phi;}}_m}{\mathrm{\&Sigma;}}{\left|{\hat{h}}_{k,j}\right|}^2+{\mathrm{\&sigma;}}_n^2}---\left(c\right)$

In formula,

represent Gaussian noise power.

The channel capacity of Cell Edge User k is:

C _k=log ₂(1+SINR _k) (d)

Here to the object of collaboration set optimization, be to make the throughput of each Cell Edge User reach maximum, that is:

$C_{\max }=\underset{k=1}{\overset{{\mathrm{UE}}_{\mathrm{CoMP}}}{\mathrm{\&Sigma;}}}\max \left(C_k\right)=\underset{k=1}{\overset{{\mathrm{UE}}_{\mathrm{CoMP}}}{\mathrm{\&Sigma;}}}\max \left({\log }_2(1+{\mathrm{SINR}}_k)\right)---\left(e\right)$

Because the channel capacity C of user k _kalong with SINR _kincrease and increase, the SINR that formula (e) is equivalent to Cell Edge User reaches maximum,

Collaboration set division methods based on UE can make the SINR sum of Cell Edge User reach maximum, but collaboration set between different user is overlapping, can make like this complexity of wireless resource scheduling very high, thereby make this method, is infeasible in actual applications.In the collaboration set based on Cell is divided, the division of static collaboration set is the collaboration set that is divided into that contiguous community is fixed, according to the concrete needs of Cell Edge User, be not optimized, so the SINR sum of the Cell Edge User in its collaboration set can be much smaller than

the shortcoming of dividing in order to overcome static collaboration set, has proposed a kind of suboptimal solution of maximization problems above here:

$\underset{k=1}{\overset{{\mathrm{UE}}_{\mathrm{CoMP}}}{\mathrm{\&Sigma;}}}\max \left({\mathrm{SINR}}_k\right)=\underset{m=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{I_m}{\mathrm{\&Sigma;}}}\max \left({\mathrm{SINR}}_{k,m}\right)\&ap;\underset{m=1}{\overset{M}{\mathrm{\&Sigma;}}}\max \left(\underset{k=1}{\overset{I_m}{\mathrm{\&Sigma;}}}{\mathrm{SINR}}_{k,m}\right)---\left(f\right)$

Even collaboration set φ _mit is maximum that the SINR sum of interior Cell Edge User reaches, and collaboration set φ _mthe SINR sum of Cell Edge User be:

$V_m=\underset{i=1}{\overset{I_m}{\mathrm{\&Sigma;}}}{\mathrm{SINR}}_i=\underset{k=1}{\overset{I_m}{\mathrm{\&Sigma;}}}\frac{\underset{i\&Element;{\mathrm{\&phi;}}_m}{\mathrm{\&Sigma;}}{\left|{\hat{h}}_{k,i}\right|}^2}{\underset{j\&NotElement;{\mathrm{\&phi;}}_m}{\mathrm{\&Sigma;}}{\left|{\hat{h}}_{k,j}\right|}^2+{\mathrm{\&sigma;}}_n^2}---\left(g\right)$

The all communities that have neither part nor lot in cooperation in system of setting up departments form set omega, and in set omega, all possible collaboration set forms set

wherein

represent k standby collaboration set in set Ψ, the c here _{i, k}for standby collaboration set

in i community.In order to overcome the shortcoming of static collaboration set division methods, a kind of new dynamic cooperative collection division methods is proposed herein.Here use V _kcollaboration set in support

weights, and select the standby collaboration set of weights maximum as a collaboration set from set Ψ, the SINR sum of Cell Edge User in collaboration set is maximized, thereby the division of collaboration set is optimized.

Based on above principle, step 3) concrete implementing procedure figure as shown in Figure 2, it specifically comprises the following steps:

3.1) establishing the set that in whole cellular mobile communication networks, collaboration set forms is Ф, and now Ф is empty set.

3.2) in cellular mobile communication networks, all communities that have neither part nor lot in cooperation form set omega, and in set omega, all possible collaboration set forms set

wherein

represent k standby collaboration set in Ψ, here c _{i, k}for collaboration set

i community, N _crepresent the number of cells in each collaboration set, then utilize formula (1) to calculate the weights of all standby collaboration sets, collaboration set

weights V _kbe expressed as:

In formula, I _krepresent collaboration set interior Cell Edge User sum, represent Gaussian noise power,

represent collaboration set

interior cell i is to the large scale decline of Cell Edge User k.

3.3) from set

middle selection collaboration set weights V _kmaximum collaboration set

if set Ф all collaboration set Zhong community all with collaboration set

zhong community is different, just by standby collaboration set

put under in set Ф;

3.4), finally order set

return step 3.2), operation is as above carried out in all communities in Ω, until the number of cells in Ω is zero; By the division of dynamic cooperative collection, be divided into altogether M collaboration set, M collaboration set forms set Ф={ φ ₁, φ ₂..., φ _m..., φ _m, φ wherein _mbe m collaboration set.

In step 3) in, user under collaboration situation, collaboration set φ _mthe reception signal indication of interior k collaboration user is:

$y_{k,m}=\underset{m\&Element;{\mathrm{\&phi;}}_k}{\mathrm{\&Sigma;}}\sqrt{p_m}{H}_{k,m}{W}_{k,m}{s}_{k,m}+\underset{n\&NotElement;{\mathrm{\&phi;}}_k}{\mathrm{\&Sigma;}}\underset{w=1}{\overset{N_U}{\mathrm{\&Sigma;}}}\sqrt{p_n}{H}_{k,n}{W}_{w,n}{s}_{w,n}+B---\left(2\right)$

In formula, y _{k, m}represent collaboration set φ _mthe reception signal of interior k Cell Edge User, p _mfor the transmitting power of community m to user k, s _{k, m}for the transmitted signal of community m to user k, H _{k, m}that community m is to the channel matrix of user k, W _{k, m}be community m to the pre-coding matrix of user k, B is Gaussian noise, N _ufor the number of each community user, in formula, first is Serving cell and the cooperation cell useful signal to user k transmission, and second is the interference to this user of other community outside collaboration set.

Fig. 3 is a kind of result that collaboration set is dynamically divided.In cellular mobile communication networks, user's channel condition information is real-time change, the large scale decline of same large channel also can be in time passing and slowly change, therefore each certain periodic feedback is fed back each user's channel condition information, and according to the result of feedback, add up each user's large scale decline, and according to the variation of large scale decline, re-start the division of collaboration set, to adapt to the now demand of Cell Edge User.

Because more slowly, therefore the feedback cycle is here long, is generally 200 to 1000 Transmission Time Intervals (TTI), thereby can effectively reduces the feedback overhead of system during the variation of the large scale of channel decline.

4), finally determine collaboration user

Calculate the Signal to Interference plus Noise Ratio SINR after each collaboration user cooperation, select new Signal to Interference plus Noise Ratio thresholding, put the user of rear Signal to Interference plus Noise Ratio SINR lower than new Signal to Interference plus Noise Ratio thresholding that cooperate under non-collaboration user, finally collaboration user number is controlled to 10% to 20%.

The Signal to Interference plus Noise Ratio SINR of collaboration set m small area edge customer k now _kcomputing formula as follows:

${\mathrm{SINR}}_k=\frac{\underset{i\&Element;{\mathrm{\&phi;}}_m}{\mathrm{\&Sigma;}}{\left|{\hat{h}}_{k,i}\right|}^2}{\underset{j\&NotElement;{\mathrm{\&phi;}}_m}{\mathrm{\&Sigma;}}{\left|{\hat{h}}_{k,j}\right|}^2+{\mathrm{\&sigma;}}_n^2}$

In formula,

represent that cell i is to the channel large scale decline of user k, represent Gaussian noise power.

Here step 2) and step 4) necessity be: in step 2) in make the user that SINR is lower put collaboration user under, after collaboration set is determined, in step 4) if in partial cooperative user Signal to Interference plus Noise Ratio gain less, illustrate that collaboration set is not too suitable to these users, in order to make Radio Resource have higher utilance, then these less users of Signal to Interference plus Noise Ratio gain are drawn in non-collaboration user.

Claims (2)

1. a division methods for dynamic cooperative collection in CoMP technology, is characterized in that, comprises the following steps:

1) set up system model

Be located in cellular mobile communication networks, have B base station, 3 * B community, each community is uniform-distribution with N _uindividual user;

2) by threshold method, determine collaboration user

Set Signal to Interference plus Noise Ratio thresholding SINR _thresholdif user's Signal to Interference plus Noise Ratio is less than SINR _threshold, put user under collaboration user;

3) determine collaboration set

(3.1) collaboration set of establishing in whole cellular mobile communication networks forms set Ф, and now Ф is empty set;

(3.2) in cellular mobile communication networks, all communities that have neither part nor lot in cooperation form set omega, and in set omega, all possible collaboration set forms set

wherein

represent k standby collaboration set in Ψ, here c _{i, k}for standby collaboration set

in i community, N _crepresent the number of cells in each collaboration set, then utilize formula (1) to calculate the weights of all standby collaboration sets, collaboration set weights V _kbe expressed as:

In formula, I _krepresent collaboration set

interior Cell Edge User sum, represent Gaussian noise power,

represent collaboration set

interior cell i is to the large scale decline of Cell Edge User k;

(3.3) from set

middle selection collaboration set weights V _kmaximum collaboration set

if set Ф all collaboration set Zhong community all with collaboration set zhong community is different, just by standby collaboration set

put under in set Ф;

(3.4) finally order set return step (3.2), operation is as above carried out in all communities in Ω, until the number of cells in Ω is zero; By the division of dynamic cooperative collection, be divided into altogether M collaboration set, M collaboration set forms set Ф={ φ ₁, φ ₂..., φ _m..., φ _m, φ wherein _mbe m collaboration set;

channel condition information by long-time statistical terminal feedback to base station obtains, and according to

variation collaboration set is revised dynamically, to meet the demand of Cell Edge User now;

4) finally determine collaboration user

Calculate the Signal to Interference plus Noise Ratio SINR of each collaboration user, select new Signal to Interference plus Noise Ratio thresholding, the user that SINR is lower than new Signal to Interference plus Noise Ratio thresholding puts non-collaboration user under, finally collaboration user number is controlled to 10% to 20%.

2. the division methods of dynamic cooperative collection in CoMP technology according to claim 1, is characterized in that step 3) in, user under collaboration situation, collaboration set φ _mthe reception signal indication of interior k collaboration user is:

$y_{k,m}=\underset{m\&Element;{\mathrm{\&phi;}}_k}{\mathrm{\&Sigma;}}\sqrt{p_m}{H}_{k,m}{W}_{k,m}{s}_{k,m}+\underset{n\&NotElement;{\mathrm{\&phi;}}_k}{\mathrm{\&Sigma;}}\underset{w=1}{\overset{N_U}{\mathrm{\&Sigma;}}}\sqrt{p_n}{H}_{k,n}{W}_{w,n}{s}_{w,n}+B---\left(2\right)$

In formula, y _{k, m}represent collaboration set φ _mthe reception signal of interior k Cell Edge User, p _mfor the transmitting power of community m to user k, s _{k, m}for the transmitted signal of community m to user k, H _{k, m}that community m is to the channel matrix of user k, W _{k, m}be community m to the pre-coding matrix of user k, B is Gaussian noise, N _ufor the number of each community user, in formula, first is Serving cell and the cooperation cell useful signal to user k transmission, and second is the interference to this user of other community outside collaboration set.

Images