CN108322271B - User-centric dynamic clustering method based on load - Google Patents

Abstract

Translated fromChinese

本发明公开了一种基于负载的以用户为中心的动态分簇方法，适用于采用多点协作传输的超密集网络。构造网络中的小基站对应的路径损耗图，对于每个用户，在路径损耗图中找出与该用户所在小小区的小基站相连的所有小基站，把这些小基站当做该用户的备选服务基站，用户根据参考信号估算备选服务基站到自身的信道衰落系数，并从备选服务基站中选出待选服务基站，每个用户将自身的待选服务基站组成的集合以及待选服务基站到自身的信道衰落系数通过其中的一个小基站发送给中心控制器，中心控制器根据接收到的信息以及每个用户的速率需求为每个用户选择服务基站簇。

The invention discloses a load-based user-centered dynamic clustering method, which is suitable for an ultra-dense network adopting multi-point cooperative transmission. Construct the path loss graph corresponding to the small base station in the network. For each user, find all the small base stations connected to the small base station in the small cell where the user is located in the path loss graph, and regard these small base stations as the user's alternative service. Base station, the user estimates the channel fading coefficient from the candidate service base station to itself according to the reference signal, and selects the candidate service base station from the candidate service base stations. The channel fading coefficient to itself is sent to the central controller through one of the small base stations, and the central controller selects a cluster of serving base stations for each user according to the received information and the rate requirement of each user.

Description

User-centered dynamic clustering method based on load

Technical Field

The invention relates to the field of communication, in particular to a load-based user-centered dynamic clustering method.

Background

Ultra Dense Networks (UDNs) are one of the key technologies for next-generation mobile communications. A large number of small base stations with low power are deployed in the UDN, so that the system capacity can be improved, the indoor coverage can be enhanced, and the frequency spectrum utilization rate can be improved. Because the distances between the small base stations are very close, if they use the same frequency band, there will be serious interference, so the effective interference management method is a hot point of research.

Although the conventional interference management strategies such as power control, scheduling based on channel quality, and the like can suppress intra-cell interference to a certain extent, interference control is performed only in a single cell, data interaction with an adjacent cell cannot be performed, and the processing capability of inter-cell interference is very limited. The Coordinated Multiple Point (CoMP) technology can avoid inter-cell interference or convert an interference link into a useful link through data interaction between base stations, and can effectively reduce interference.

CoMP is divided into coordinated scheduling/coordinated beamforming, dynamic node selection, and Joint Transmission (JT). In the cooperative scheduling/cooperative beamforming mode, a base station of a cell where a user is located provides service for the user, and a scheduling and precoding scheme of the user is jointly determined by a plurality of base stations. In the dynamic node selection mode, one of the base stations is selected to transmit data for the user. In JT mode, multiple base stations share user data, and transmit data to users at the same time, converting an interfering link into a useful link. In JT-CoMP, channel state information and user data are exchanged between base stations, which requires too high backhaul link overhead if all small base stations cooperate with each other, and cannot provide sufficient gain if there are fewer cooperating base stations, so that selection of a cooperating cluster is crucial.

Clustering methods in JT-CoMP, i.e., methods for a user to select a serving base station cluster, are classified into three types: static clustering, semi-dynamic clustering, and dynamic clustering. The static clustering method designs a fixed cooperative cluster which does not change along with time in advance according to a network structure and the geographic position of a base station, and has the advantages of simple operation and extremely low computation amount, and the defects of no consideration of the time-varying characteristic of a fading channel and limited obtainable cooperative gain. Semi-dynamic clustering is an improvement over static clustering, with higher cooperative gains than static clustering. The dynamic clustering periodically clusters according to the change of the channel state information and the user position, and can obtain higher cooperative gain, but the clustering complexity is higher. Dynamic clustering is divided into three categories: network-centric dynamic clustering, user-centric dynamic clustering, and hybrid clustering. In dynamic clustering with a network as a center, a base station is clustered at first, each base station cluster comprises a plurality of base stations, and then a proper service base station is selected for a user in the base station cluster where the user is located. The dynamic clustering taking the user as the center selects a proper service base station cluster for the user in all small base stations according to the channel state information of the user, and the method can obtain higher signal-to-interference-and-noise ratio and throughput, but has the highest complexity. Hybrid clustering is a combination of network-centric clustering and user-centric clustering that can be adjusted as network conditions change, making a trade-off between complexity and cooperative gain.

The document "Load aware self-organizing user-centralized dynamic clustering for 5G networks" proposes a Load-based user-centric dynamic clustering method that can significantly reduce the number of users that fail to meet rate requirements due to excessive Load. The document mentions: the method can also be regarded as a hybrid clustering, and comprises the steps of firstly clustering by taking a network as a center, forming a base station cluster by small base stations in the coverage range of the same macro base station, then clustering by taking users as the center in the base station cluster, and selecting a service base station cluster for each user. However, if the user is located at the boundary of the coverage area of the macro base station, the number of base stations within the coverage area of the macro base station serving the user is small, and the neighboring macro base station and the small base station within the coverage area of the macro base station and close to the user may cause serious interference to the user.

Disclosure of Invention

The invention provides a load-based user-centered dynamic clustering method, which is suitable for an ultra-dense network adopting multi-point cooperative transmission.

The technical idea for realizing the invention is as follows: constructing a path loss diagram corresponding to small base stations in a network, for each user, finding out all the small base stations connected with the small base station of the small cell where the user is located in the path loss diagram, using the small base stations as alternative service base stations of the user, estimating a channel fading coefficient from the alternative service base stations to the user by the user according to a reference signal, selecting the service base stations to be selected from the alternative service base stations, sending a set formed by the service base stations to be selected and the channel fading coefficient from the service base stations to the user to a central controller by each user through one of the small base stations, and selecting a service base station cluster for each user by the central controller according to received information and the rate requirement of each user.

In order to realize the technical idea, the load-based user-centric dynamic clustering method provided by the invention is suitable for a super-dense network adopting CoMP, and comprises the following steps:

a, constructing a path loss graph corresponding to small base stations in a network, wherein nodes in the graph correspond to the small base stations, edges correspond to path loss between the small base stations, if the path loss between the small base stations is smaller than a preset loss threshold, edges exist between the nodes corresponding to the two small base stations, and if the path loss between the small base stations is larger than the preset loss threshold, edges do not exist between the nodes corresponding to the two small base stations;

b, for the u-th user, all the small base stations connected with the small base station of the small cell where the user is located are found in the path loss diagram, and the small base stations are placed in the set Q_uWhere U is 1,2, …, U being the total number of users in the network, Q_uThe small base station in the U is an alternative service base station of the u user;

c, each small base station sends reference signals with the same power, the u-th user measures the received signal power according to the received reference signals and estimates Q_uThe channel fading coefficient from the small base station to the small base station is 1,2, …, and U is the total number of users in the network;

d, the u-th user compares the power of the received signals measured in the step C with a preset power threshold, selects the small base stations corresponding to a plurality of received signals larger than the power threshold, and sequentially puts the small base stations corresponding to the received signals in the set P according to the sequence of the power from large to small_uIn (1),

P_uelement in (1) represents a small base station, u_kIs P_uTotal number of elements in (1), u_kNot more than Q_uThe total number of middle elements, U ═ 1,2, …, U being the total number of users in the network, P_uThe small base station in the U is a service base station to be selected of the u user;

e, the u-th user will collect P_uAnd P estimated in step C_uThe channel fading coefficient from each small base station to itself is sent to the set P_uFirst small base station BS in_u,1Small base station BS_u,1Sending the received information to a central controller;

f, the central controller receives the P based on the speed requirement of each user and the step E_uHexinSelecting a service base station for each user by the channel fading coefficient;

g, the central controller is according to N_uFinding out user set V served by jth small base station_jAnd will V_jSending the information to the jth small base station, the jth small base station and the set V_jJ is 1,2, …, J is the total number of small base stations in the network.

Further, the step F specifically includes:

f1, using N_uServing base station cluster, N, representing the u-th user_uSetting the set M as an empty set, where U is 1,2, …, and U is the total number of users in the network;

f2, small base station BS_u,1Serving base station cluster N for joining to u-th user_uThe central controller calculates the rate r of the user when only 1 base station serves the u-th user at the moment according to the known channel fading coefficient_u,1U is 1,2, …, U being the total number of users in the network, r_u,1The index 1 in (a) indicates that only 1 base station serves the user, i is 2, i is the total number of base stations serving the u-th user, and i is 2 indicates the rate of the user when 2 small base stations are considered to serve the u-th user in step F3;

f3, using | P_uAn | representation set P_uPotential of (c), if P_u| ≧ i, the central controller considers the small base station BS_u,iWhen the serving base station of the U-th user, U is 1,2, …, U is not in the set M, U is the total number of users in the network, the central controller calculates the rate r of the U-th user at the moment according to the known channel fading coefficient_u,i，r_u,iThe index i in (1) indicates that there are i base stations serving the user, i ═ 1,2, …, u_k；

F4, comparison r_u,i-1、r_u,iAnd d_u，d_uIs the lowest rate requirement for the u-th user, if r_u,i-1＞r_u,iThen not sending the small base station BS_u,iServing base station cluster N for joining to u-th user_uAnd put u in set M if d_u>r_u,i>r_u,i-1Or r_u,i>d_u>r_u,i-1Then the small base station BS is set_u,iServing base station cluster N for joining to u-th user_uIf r is_u,2>r_u,1>d_uThen not sending the small base station BS_u,iServing base station cluster N for joining to u-th user_uLet i be i +1, U be 1,2, …, U being the total number of users in the network;

f5, repeating the steps F3 and F4 until i > max { | P_u| U ═ 1,2, …, U and U are not in set M }, U being the total number of users in the network.

In the invention, for each user, all small base stations connected with the small base station of the small cell where the user is located are found in a path loss diagram, the small base stations are used as alternative service base stations of the user, the small base stations are not limited in the coverage range of a single macro base station, and can be in the coverage ranges of a plurality of macro base stations, and as long as the small base stations with better channel quality with the user are used as the alternative service base stations of the user, the defects in the document 'Load aware self-organization user-center dynamic coordination for 5G networks' are overcome.

Drawings

FIG. 1 is a path loss diagram of an embodiment of the present invention;

FIG. 2 is a flow chart of the present invention;

fig. 3 is a flow chart of the present invention for selecting a serving base station cluster for each user.

Detailed Description

An embodiment of the present invention is given below, and the present invention will be described in further detail. Consider a very dense network comprising several small base stations, one in each small cell, and a number of users, both randomly distributed within the network. Each small base station is connected with the central controller through a backhaul link.

The central controller first constructs a path loss graph corresponding to small base stations in a network, as shown in fig. 1, a node in the graph corresponds to a small base station, an edge corresponds to a path loss between the small base stations, if the path loss between the small base stations is smaller than a preset loss threshold, an edge exists between nodes corresponding to the two small base stations, and if the path loss between the small base stations is smaller thanAnd if the loss is greater than the preset loss threshold, no edge exists between the nodes corresponding to the two small base stations. As an example, there are 24 small base stations in FIG. 1, the circle represents a small base station, the number in the circle represents the serial number of the small base station, and BS is used_jDenotes the j-th small base station, j is 1,2, …, 24.

For the U-th user, if the central controller selects a serving base station for the user from all the small base stations, the channel fading coefficient from each small base station to the user needs to be known, which requires a higher backhaul link, U is 1,2, …, and U is the total number of users in the network. If the distance between a small base station and the u-th user is long, and the small base station cannot become the serving base station of the u-th user, the channel fading coefficients from the small base station to the u-th user are still sent to the central controller, which causes waste of link resources and does not gain any more, and therefore, it is not necessary to use all small base stations as alternative serving base stations of the u-th user. By using

The small base station of the small cell in which the u-th user is located is indicated, in the path loss diagram, if some small base stations are connected with the small base station

Then the small base stations are connected with

The path loss between the small base stations is small, the channel quality is high, and similarly, the channel quality between the small base stations and the u-th user is also high, so that the small base stations need to be used as alternative serving base stations of the u-th user. If a small base station is located in the path loss diagram and

if not, the small base station and

the path loss between the small base station and the u-th user is large, the channel quality between the small base station and the u-th user is poor, and the small base station and the u-th user are not necessary to be usedAnd the small base station is taken as an alternative service base station of the u user.

For the u-th user, all the small base stations connected with the small base station of the small cell where the user is located are found in the path loss diagram, and the small base stations are put into a set

In (1),

the small base station in (1) is an alternative serving base station of the u-th user.

Is a set of candidate serving base stations for the u-th user. When k is_uWhen the values of (A) are 1,2, … and 24 respectively,

as follows.

Q₁＝{BS₂,BS₇,BS₈}

Q₂＝{BS₁,BS₃,BS₇,BS₁₀}

Q₃＝{BS₂,BS₄,BS₁₀}

Q₄＝{BS₃,BS₅,BS₆,BS₁₁}

Q₅＝{BS₄,BS₁₁,BS₁₂}

Q₆＝{BS₄,BS₁₀,BS₁₄}

Q₇＝{BS₁,BS₂,BS₈,BS₁₆}

Q₈＝{BS₁,BS₇,BS₉}

Q₉＝{BS₈,BS₁₆}

Q₁₀＝{BS₂,BS₃,BS₆,BS₁₄,BS₁₅,BS₁₇}

Q₁₁＝{BS₄,BS₅,BS₁₃,BS₁₄}

Q₁₂＝{BS₅,BS₁₃}

Q₁₃＝{BS₁₁,BS₁₂,BS₁₄,BS₁₈,BS₁₉}

Q₁₄＝{BS₆,BS₁₀,BS₁₁,BS₁₃,BS₁₇,BS₁₈,BS₂₁}

Q₁₅＝{BS₂,BS₁₀,BS₁₇,BS₁₆,BS₂₂}

Q₁₆＝{BS₇,BS₉,BS₁₅,BS₂₂}

Q₁₇＝{BS₁₀,BS₁₄,BS₁₅,BS₂₂,BS₂₄}

Q₁₈＝{BS₁₃,BS₁₄,BS₁₉,BS₂₀}

Q₁₉＝{BS₁₃,BS₁₈,BS₂₀}

Q₂₀＝{BS₁₈,BS₁₉,BS₂₁,BS₂₄}

Q₂₁＝{BS₁₄,BS₂₀,BS₂₄}

Q₂₂＝{BS₁₅,BS₁₆,BS₁₇,BS₂₃,BS₂₄}

Q₂₃＝{BS₂₂,BS₂₄}

Q₂₄＝{BS₁₇,BS₂₀,BS₂₁,BS₂₂,BS₂₃}

Each small base station transmits reference signals with the same power, and the u-th user measures the received signal power according to the received reference signals and estimates

And U is the total number of users in the network. For the u th useComparing the measured powers of multiple received signals with a preset power threshold by a user, selecting small base stations corresponding to the multiple received signals larger than the power threshold, and sequentially placing the small base stations corresponding to the multiple received signals in a set P according to the sequence of the powers from large to small_uIn (1),

P_uelement in (1) represents a small base station, u_kIs P_uThe total number of elements in (c). Set P_uThe small base station in (1) and the u-th user have better channel quality and are the candidate service base stations of the u-th user.

The known set P is required for the central controller to select a serving base station cluster for each user_uAnd set P_uThe channel fading coefficient from the small base station to the u-th user. The u-th user will set P_uAnd estimating the resulting P_uSending the channel fading coefficient from each small base station to itself to the small base station BS_u,1Small base station BS_u,1The received information is sent to the central controller.

Central controller based on each user's rate requirement, set P_uAnd set P_uThe channel fading coefficients from the small base station to the U-th user in the network are used for selecting a serving base station cluster for each user, where U is 1,2, …, and U is the total number of users in the network, and the specific process is as follows:

step 1, with N_uServing base station cluster, N, representing the u-th user_uSetting the set M as an empty set, where U is 1,2, …, and U is the total number of users in the network;

step 2, the small base station BS_u,1Serving base station cluster N for joining to u-th user_uThe central controller calculates the rate r of the current u-th user according to the known channel fading coefficient_u,1U is 1,2, …, U is the total number of users in the network, let i be 2;

step 3, using | P_uAn | representation set P_uPotential of (c), if P_u| ≧ i, the central controller considers the small base station BS_u,iAs the serving base station of the U-th user, U is 1,2, …, U is not presentIn the set M, U is the total number of users in the network, and the central controller calculates the rate r of the current U-th user according to the known channel fading coefficient_u,i；

Step 4, comparing r_u,i-1、r_u,iAnd d_u，d_uIs the lowest rate requirement for the u-th user, if r_u,i-1＞r_u,iThen not sending the small base station BS_u,iServing base station cluster N for joining to u-th user_uAnd put u in set M if d_u>r_u,i>r_u,i-1Or r_u,i>d_u>r_u,i-1Then the small base station BS is set_u,iServing base station cluster N for joining to u-th user_uIf r is_u,2>r_u,1>d_uThen not sending the small base station BS_u,iServing base station cluster N for joining to u-th user_uLet i be i +1, U be 1,2, …, U being the total number of users in the network;

step 5, repeating the steps F3 and F4 until i > max { | P_u| U ═ 1,2, …, U and U are not in set M }, U being the total number of users in the network.

The central controller is based on N_uFinding out user set V served by jth small base station_jAnd will V_jSending the information to the jth small base station, the jth small base station and the set V_jJ ═ 1,2, …, 24.

With reference to the flowchart of the present invention, i.e., fig. 2, the specific steps of the user-centric dynamic clustering method based on load are as follows:

a, constructing a path loss graph corresponding to small base stations in a network, wherein nodes in the graph correspond to the small base stations, edges correspond to path loss between the small base stations, if the path loss between the small base stations is smaller than a preset loss threshold, edges exist between the nodes corresponding to the two small base stations, and if the path loss between the small base stations is larger than the preset loss threshold, edges do not exist between the nodes corresponding to the two small base stations;

b, for the u-th user, all the small base stations connected with the small base station of the small cell where the user is located are found in the path loss diagram, and the small base stations are placed in the small base stationsSet Q_uWhere U is 1,2, …, U being the total number of users in the network, Q_uThe small base station in the U is an alternative service base station of the u user;

c, each small base station sends reference signals with the same power, the u-th user measures the received signal power according to the received reference signals and estimates Q_uThe channel fading coefficient from the small base station to the small base station is 1,2, …, and U is the total number of users in the network;

d, the u-th user compares the power of the received signals measured in the step C with a preset power threshold, selects the small base stations corresponding to a plurality of received signals larger than the power threshold, and sequentially puts the small base stations corresponding to the received signals in the set P according to the sequence of the power from large to small_uIn (1),

P_uelement in (1) represents a small base station, u_kIs P_uTotal number of elements in (1), u_kNot more than Q_uThe total number of middle elements, U ═ 1,2, …, U being the total number of users in the network, P_uThe small base station in the U is a service base station to be selected of the u user;

e, the u-th user will collect P_uAnd P estimated in step C_uSending the channel fading coefficient from each small base station to itself to the small base station BS_u,1Small base station BS_u,1Sending the received information to a central controller;

f, the central controller receives the P based on the speed requirement of each user and the step E_uAnd selecting a serving base station for each user by the channel fading coefficient;

g, the central controller is according to N_uFinding out user set V served by jth small base station_jAnd will V_jSending the information to the jth small base station, the jth small base station and the set V_jJ is 1,2, …, J is the total number of small base stations in the network.

With reference to the flowchart of the present invention, i.e. fig. 3, for selecting a serving base station cluster for each user, the specific steps of selecting a serving base station cluster for each user are as follows:

f1, using N_uServing base station cluster, N, representing the u-th user_uSetting the set M as an empty set, where U is 1,2, …, and U is the total number of users in the network;

f2, small base station BS_u,1Serving base station cluster N for joining to u-th user_uThe central controller calculates the rate r of the current u-th user according to the known channel fading coefficient_u,1U is 1,2, …, U is the total number of users in the network, let i be 2;

f3, using | P_uAn | representation set P_uPotential of (c), if P_u| ≧ i, the central controller considers the small base station BS_u,iWhen the serving base station of the U-th user, U is 1,2, …, U is not in the set M, U is the total number of users in the network, the central controller calculates the rate r of the U-th user at the moment according to the known channel fading coefficient_u,i；

F4, comparison r_u,i-1、r_u,iAnd d_u，d_uIs the lowest rate requirement for the u-th user, if r_u,i-1＞r_u,iThen not sending the small base station BS_u,iServing base station cluster N for joining to u-th user_uAnd put u in set M if d_u>r_u,i>r_u,i-1Or r_u,i>d_u>r_u,i-1Then the small base station BS is set_u,iServing base station cluster N for joining to u-th user_uIf r is_u,2>r_u,1>d_uThen not sending the small base station BS_u,iServing base station cluster N for joining to u-th user_uLet i be i +1, U be 1,2, …, U being the total number of users in the network;

f5, repeating the steps F3 and F4 until i > max { | P_u| U ═ 1,2, …, U and U are not in set M }, U being the total number of users in the network.

The above embodiments are merely illustrative of the present invention, and those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.

Claims (1)

Translated fromChinese

1.一种基于负载的以用户为中心的动态分簇方法，适用采用多点协作传输的超密集网络，其特征在于：包括如下步骤：1. a user-centered dynamic clustering method based on load, suitable for adopting the ultra-dense network of coordinated multi-point transmission, is characterized in that: comprise the steps:A，构造网络中的小基站对应的路径损耗图，该图中的节点对应小基站，边对应小基站之间的路径损耗，若小基站之间的路径损耗小于预先设置的损耗门限，则这两个小基站对应的节点之间有条边，若小基站之间的路径损耗大于预先设置的损耗门限，则这两个小基站对应的节点之间没有边；A. Construct the path loss graph corresponding to the small base stations in the network. The nodes in the graph correspond to the small base stations, and the edges correspond to the path loss between the small base stations. If the path loss between the small base stations is less than the preset loss threshold, then this There is an edge between the nodes corresponding to the two small base stations. If the path loss between the small base stations is greater than the preset loss threshold, there is no edge between the nodes corresponding to the two small base stations;B，对于第u个用户，在路径损耗图中找出与该用户所在小小区的小基站相连的所有小基站，将这些小基站放在集合Q_u中，u＝1,2,…,U，U是该网络中用户的总数，Q_u中的小基站是第u个用户的备选服务基站；B. For the uth user, find all the small base stations connected to the small base station of the small cell where the user is located in the path loss graph, and put these small base stations in the set Qu,_u =1,2,...,U , U is the total number of users in the network, and the small base station in Qu is the candidate serving base station of the_uth user;C，每个小基站以相同的功率发送参考信号，第u个用户根据接收到的参考信号测量接收信号功率并且估算Q_u中的小基站到自身的信道衰落系数，u＝1,2,…,U，U是该网络中用户的总数；C, each small base station transmits a reference signal with the same power, the uth user measures the received signal power according to the received reference signal and estimates the channel fading coefficient from the small base station in Qu to itself,_u =1,2,... , U, U is the total number of users in the network;D，第u个用户将步骤C测量得到的多个接收信号功率与预先设置的功率门限相比较，选出大于功率门限的若干个接收信号对应的小基站，并且按功率从大到小的顺序将它们对应的小基站依次放在集合P_u中，

P_u中的元素表示小基站，u_k是P_u中元素的总数，u_k不大于Q_u中元素的总数，u＝1,2,…,U，U是该网络中用户的总数，P_u中的小基站是第u个用户的待选服务基站；D, the u-th user compares the multiple received signal powers measured in step C with the preset power threshold, selects the small base stations corresponding to several received signals greater than the power threshold, and arranges the power in descending order Put their corresponding small base stations in the set P_u in turn,

The elements in P_u represent small base stations,_uk is the total number of elements in P_u , u_k is not greater than the total number of elements in Q_u , u=1,2,…,U, U is the total number of users in the network, P The small base station in_u is the candidate serving base station of the u-th user;E，第u个用户将集合P_u以及步骤C估算得到的P_u中的每个小基站到自身的信道衰落系数发送给集合P_u中的第一个小基站BS_u,1，小基站BS_u,1将收到的这些信息发送给中心控制器；E, the uth user sends the set P_u and the channel fading coefficient from each small base station in P_u estimated in step C to itself to the first small base station BS_u,1 in the set P_u , and the small base station BS_u,1 sends the received information to the central controller;F，中心控制器基于每个用户的速率需求、步骤E接收到的P_u和信道衰落系数为每个用户选择服务基站，具体过程如下：F. The central controller selects a serving base station for each user based on the rate requirement of each user, the P_u received in step E, and the channel fading coefficient. The specific process is as follows:F1，用N_u表示第u个用户的服务基站簇，N_u为空集，令集合M为空集，u＝1,2,…,U，U是该网络中用户的总数；F1, use Nu to represent the serving base station cluster of the_uth user, Nu_is an empty set, let the set M be an empty set, u=1,2,...,U, U is the total number of users in the network;F2，将小基站BS_u,1加入到第u个用户的服务基站簇N_u，中心控制器根据已知的信道衰落系数计算此时只有1个基站为第u个用户服务时该用户的速率r_u,1，u＝1,2,…,U，U是该网络中用户的总数，r_u,1中的下标1表示只有1个基站为该用户服务，令i＝2，i是为第u个用户服务的基站的总数，i＝2表示在步骤F3中将考虑2个小基站为第u个用户服务时该用户的速率；F2, add the small base station BS_u,1 to the serving base station cluster Nu of the_u -th user, and the central controller calculates the rate of the user when there is only one base station serving the u-th user according to the known channel fading coefficient. r_u,1 , u=1,2,...,U, U is the total number of users in the network, the subscript 1 in r_u,1 indicates that only one base station serves the user, let i=2, i is The total number of base stations serving the u-th user, i=2 represents the rate of the user when two small base stations serve the u-th user in step F3;F3，用|P_u︱表示集合P_u的势，若|P_u︱≥i，中心控制器考虑将小基站BS_u,i当做第u个用户的服务基站，u＝1,2,…,U且u不在集合M中，U是该网络中用户的总数，中心控制器根据已知的信道衰落系数计算此时第u个用户的速率r_u,i，r_u,i中的下标i表示有i个基站为该用户服务，i＝1,2,…,u_k；F3, use |P_u ︱ to represent the potential of the set P_u , if |P_u ︱ ≥ i, the central controller considers the small base station BS_u,i as the serving base station of the u-th user, u=1,2,…, U and u is not in the set M, U is the total number of users in the network, the central controller calculates the rate r_u,i of the u-th user at this time according to the known channel fading coefficient, and the subscript i in r_u, i Indicates that there are i base stations serving the user, i=1,2,...,u_k ;F4，比较r_u,i-1、r_u,i和d_u，d_u是第u个用户的最低速率需求，若r_u,i-1＞r_u,i，则不把小基站BS_u,i加入到第u个用户的服务基站簇N_u且将u放入集合M中，若d_u>r_u,i>r_u,i-1或r_u,i>d_u>r_u,i-1，则将小基站BS_u,i加入到第u个用户的服务基站簇N_u，若r_u,2>r_u,1>d_u，则不把小基站BS_u,i加入到第u个用户的服务基站簇N_u，令i＝i+1，u＝1,2,…,U，U是该网络中用户的总数；F4, compare r_u,i-1 , r_u,i and_du ,_du is the minimum rate requirement of the u-th user, if r_u,i-1 >r_u,i , the small base station BS_u is not_{, i} joins the u-th user's serving base station cluster Nu and puts_u into the set M, if d_u >r_u,i >r_u,i-1 or r_u,i >d_u >r_{u, i-1} , then add the small base station BS_u,i to the serving base station cluster Nu of the_u -th user, if r_u,2 >r_u,1 >d_u , then do not add the small base station BS_u,i to The serving base station cluster Nu of the_uth user, let i=i+1, u=1,2,...,U, U is the total number of users in the network;F5，重复步骤F3和F4，直至i＞max{|P_u︱,u＝1,2,…,U且u不在集合M中}，U是该网络中用户的总数；F5, repeat steps F3 and F4 until i>max{|P u︱,_u =1,2,...,U and u is not in the set M}, U is the total number of users in the network;G，中心控制器根据N_u找到第j个小基站服务的用户集合V_j，并将V_j发送给第j个小基站，第j个小基站与集合V_j中的用户交互数据，j＝1,2,…,J，J是该网络中小基站的总数。G, the central controller finds the user set_Vj served by the j th small base station according to Nu , and sends V_j to the j th small base station. The j th small base station interacts with the users in the set V_j , j= 1,2,…,J, J is the total number of small base stations in the network.

Images