CN109831794B - Base station clustering method based on density and minimum distance in ultra-dense network - Google Patents
Base station clustering method based on density and minimum distance in ultra-dense network Download PDFInfo
- Publication number
- CN109831794B CN109831794B CN201910222536.0A CN201910222536A CN109831794B CN 109831794 B CN109831794 B CN 109831794B CN 201910222536 A CN201910222536 A CN 201910222536A CN 109831794 B CN109831794 B CN 109831794B
- Authority
- CN
- China
- Prior art keywords
- base station
- density
- micro cell
- cluster center
- cell base
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention discloses a base station clustering method based on density and minimum distance in an ultra-dense network, which comprises the steps of firstly calculating the distribution density and clustering density threshold of each micro-cell base station in the ultra-dense network, and enabling the micro-cell base stations with the distribution density larger than the clustering density threshold to form an initial cluster center pool; calculating the minimum value of the distance between each micro cell base station in the initial cluster center pool and the micro cell base station higher than the distribution density of the micro cell base stations, defining the product of the distribution density of the micro cell base stations and the minimum distance as a weighted distribution density, and obtaining the cluster center pool to be selected according to the weighted distribution density; calculating cluster center isolation distances, and removing cluster centers with smaller weighted distribution density values from two cluster centers with distances between every two cluster centers which are greater than the cluster center isolation distance in the cluster center pool to be selected from the cluster center pool to be selected in sequence; finally, the cluster center number in the cluster center pool to be selected and the geographic position of the cluster center base station are used as parameters of a traditional K-means algorithm, and the K-means algorithm is executed to obtain a clustering result; the invention solves the problem of non-uniform clustering.
Description
Technical Field
The invention belongs to the technical field of wireless communication, is applied to the management of microcell base stations in an ultra-dense network, and particularly relates to a base station clustering method based on density and minimum distance in the ultra-dense network.
Background
With the rapid development of mobile communication technology, communication devices connected to a network are growing explosively, so that the originally huge network structure becomes more complex. At the same time, the demand of users for data traffic is also increasing dramatically, and wireless networks will face tremendous pressure and challenges. In an ultra-dense network, the microcell base stations are densely deployed in the cells, so that the system capacity of the wireless network can be greatly improved. However, due to the intensive deployment of a large number of base stations, the interference problem in the network is also becoming more serious, and the problem of unreasonable resource allocation is urgently to be solved. In the ultra-dense network, the whole network is divided into a plurality of sub-networks by the clustering technology, and then interference management and resource allocation are carried out in each sub-network, so that the network topology structure can be effectively simplified, the base station can be managed conveniently, and therefore interference management and resource allocation are effectively carried out. The base stations can be clustered effectively and rapidly according to the positions of the microcell base stations by adopting a K-means algorithm. However, the traditional K-means algorithm needs to manually set the number of clusters, which results in that the algorithm cannot adapt to the change of the network topology. In addition, the algorithm randomly selects the initial cluster center, so that the final clustering result is easy to fall into a local optimal solution.
Therefore, a clustering method capable of adapting to a changing network topology and better clustering the network is needed to be found, so that the method can be practically applied to the ultra-dense network.
Disclosure of Invention
The invention aims to provide a base station clustering method based on density and minimum distance in an ultra-dense network, which can dynamically cluster massive base stations according to the change of network topology, avoid the situation of falling into a local optimal solution by screening cluster center points, improve the accuracy of clustering, accelerate the convergence speed of clustering, solve the problem of non-uniform clustering result in the traditional scheme, and is suitable for the efficient management of the base stations in the ultra-dense network, and the specific technical scheme is as follows:
a method for clustering base stations based on density and minimum distance in an ultra-dense network, the method comprising the steps of:
s1, recording the geographical positions of N micro cell base stations in the ultra-dense network, and calculating the Euclidean distance between any two micro cell base stations;
s2, calculating the distribution density and clustering density threshold of each micro cell base station in the ultra-dense network, comparing the distribution density of each micro cell base station with the clustering density threshold, and taking the micro cell base station with the distribution density larger than the clustering density threshold as an initial cluster center;
s3, forming an initial cluster center pool by all the initial cluster centers, and calculating the minimum value of the distance between each micro cell base station in the initial cluster center pool and the micro cell base station with higher distribution density;
s4, calculating the product of the distribution density of each micro cell base station in the initial cluster center pool and the minimum distance value between the micro cell base stations higher than the distribution density, and recording the product as weighted distribution density thetajAnd corresponding the micro cell base station to the weighted distribution density thetajThe cluster center pools to be selected are formed by descending order from large to small;
s5, calculating the cluster center separation distance according toAccording to the weighted distribution density thetajThe cluster centers with smaller weighted distribution density value in the two cluster centers with the distance between every two cluster centers smaller than the cluster center isolation distance in the cluster center pool to be selected are sequentially removed from the cluster center pool to be selected;
s6, counting and recording the number K of cluster centers in the cluster center pool to be selected and the geographical position of each cluster center, inputting the number K and the geographical position of each cluster center as parameters into a traditional K-means algorithm, and executing the K-means algorithm to obtain clustering results of all micro cell base stations in the ultra-dense network.
Further, in step S2, the distribution density is defined as:
the clustering density threshold is defined as:
Further, in step S3, the larger the minimum value is, the more uniform the distribution of the microcell picocell base stations in the initial cluster center is, and the minimum value of the distance between the microcell base stations with the maximum distribution density in the initial cluster center is the distance between the microcell base stations with the farthest distance therefrom;
and all the initial cluster centers in the initial cluster center pool are arranged in a descending order according to the distribution density of the micro cell base stations.
Further, in step S4, the weighted distribution density θjThe greater the description relates to the weighted distribution density θjThe larger the distribution density of the corresponding micro cell base station is, and the more the distribution density theta is corresponding to the weight distribution densityjThe farther the corresponding microcell base station is from the other microcell base stations.
Further, in step S5, the cluster center separation distance is defined as:
Further, in step S5, the euclidean distance d between any microcell base station i and any microcell base station j in the ultra-dense network is calculatedi,jIs separated from the cluster center by a distance RcComparison, if di，j＜RcRemoving the micro cell base station j from the center pool of the cluster to be selected until the distance d between any two micro cell base stations in the center pool of the cluster to be selectedi,jAre all greater than the cluster center separation distance Rc。
The invention relates to a base station clustering method based on density and minimum distance in an ultra-dense network, which comprises the steps of firstly calculating the distribution density and clustering density threshold of microcell base stations in the ultra-dense network, constructing an initial cluster center pool for the microcell base stations with the distribution density larger than the clustering density threshold, and calculating the distance between any two microcell base stations in the initial cluster center pool; then, calculating the weighted distribution density of each micro cell base station in the initial cluster center pool as the product of the actual distribution density and the minimum distance from the high-density base station, and arranging the base stations in the pool in a descending order according to the weighted distribution density to form a center pool to be selected; calculating the cluster center isolation distance of the pool to be selected; comparing the distance between any two micro cell base stations in the initial cluster center pool with the cluster center isolation distance, and removing the micro cell base stations smaller than the cluster center isolation distance from the cluster center pool to be selected; finally, counting and recording the number and the geographic position of the micro cell base stations in the central pool of the cluster to be selected, inputting the number and the geographic position as parameters of the traditional K-means, and executing a K-means algorithm to obtain a clustering result of the micro cell base stations in the ultra-dense network; compared with the prior art, the invention can adaptively set the number of clusters according to the position change of the base station in the network, thereby better realizing the clustering of the ultra-dense network; the initial cluster centers are jointly screened by calculating the distribution density and the minimum distance of the base station, so that the situation of falling into a local optimal solution can be avoided, and a clustering result with higher accuracy is obtained.
Drawings
Fig. 1 is a block diagram illustrating a flow of a base station clustering method based on density and minimum distance in an ultra-dense network according to an embodiment of the present invention.
Fig. 2 is a diagram illustrating a distribution position simulation of a femtocell base station according to an embodiment of the present invention.
Fig. 3 is a diagram of clustering results according to an embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention.
In the embodiment of the invention, a base station clustering method based on density and minimum distance in an ultra-dense network is provided; specifically, the method of the present invention is described by performing a distribution position simulation of the micro cell base stations in a 300m by 300m area, where the position distribution of all the micro cell base stations in the area satisfies an independent poisson point distribution process; referring to fig. 1, assuming that the number of microcell base stations in an area is N equal to 50, a specific process of the method includes the steps of:
step one, recording the geographical positions of N microcell base stations in the ultra-dense network, that is, N is 50, referring to fig. 2, which is a distribution position simulation diagram of the microcell base stations in this embodiment, from which the actual position distribution situation of each microcell base station in the ultra-dense network can be known, so that the geographical position of each microcell base station can be obtained; then calculating the Euclidean distance between any two microcell base stations; in the super-dense network, the base station i and the base station j are both 2-dimensional vectors, and the euclidean distance between the base station i and the base station j is calculated as follows:
Step two,Calculating the distribution density and clustering density threshold of each micro cell base station in the ultra-dense network, wherein the distribution density is defined as:
in the formula (d)i,jRepresenting the Euclidean distance between a microcell base station i and a microcell base station j in the ultra-dense network; the clustering density threshold is defined as:
in the formula, alpha is a control coefficient of the number of the initial cluster centers, and the value range is alpha belongs to [0.5,1 ]]And the larger α is, the less the number of initial cluster centers contained in the super-dense network is, and this embodiment specifically describes the method of the present invention with α being 1; comparing the obtained distribution density of each micro cell base station with the clustering density threshold value, and if the distribution density corresponding to the micro cell base station is greater than the clustering density threshold value, namely
The corresponding microcell base station j is taken as the initial cluster center.
Step three, forming an initial cluster center pool by all initial cluster centers, and calculating the minimum value of the distance between each micro cell base station in the initial cluster center pool and a micro cell base station with higher distribution density; in the embodiment, all the initial cluster centers in the initial cluster center pool are arranged in a descending order according to the distribution density of the corresponding micro cell base stations; the larger the corresponding minimum value of the designated micro cell base station is, the more uniform the distribution of the micro cell micro base stations in the initial cluster center is, and the minimum value of the distance between the micro cell base stations with the maximum distribution density in the initial cluster center is the distance between the micro cell base stations with the farthest distance.
Step four, calculating the product of the distribution density of all the microcell base stations in the initial cluster center pool and the minimum distance value between the microcell base stations higher than the distribution density, and recording the product as weighted distribution density thetajAnd correspondingly weighting the microcell base station to obtain the distribution density thetajThe cluster center pools to be selected are formed by descending order from large to small; specifically, the microcell base stations are j toThe minimum value among the distances of base stations having a greater distribution density (i.e., high-density base stations) is defined as δjThen the weight density θjCan be positioned as follows: thetaj＝δj·ρj(ii) a In a particular embodiment, the distribution density θ is weightedjThe larger the description and weighting distribution density θjThe greater the distribution density of the corresponding microcell base station, and the greater the weighted distribution density thetajThe farther the corresponding microcell base station is from the other microcell base stations.
Step five, calculating the cluster center isolation distance of the cluster center pool to be selected, and sequentially removing the cluster centers with smaller weighted distribution density values from the two cluster centers with the distances between every two cluster centers smaller than the cluster center isolation distance in the cluster center pool to be selected from the cluster center pool to be selected according to the front-back arrangement sequence of the cluster centers in the cluster center pool to be selected; in the embodiment of the present invention, the cluster center isolation distance is defined as:
in the formula, beta is a control coefficient of the number of the centers of the initial clusters, and the value range is beta from [0.5,1 ]]Wherein, if β is smaller, it means that the cluster center separation distance is smaller.
Obtaining the cluster center isolation distance R in the cluster center pool to be selectedcThen, the Euclidean distance d between any micro cell base station i and any micro cell base station j in the ultra-dense network is determinedi,jDistance R from cluster centercComparison, if di，j＜RcThen the micro cell base station j is removed from the center pool of the cluster to be selected until the distance d between any two micro cell base stations in the center pool of the cluster to be selectedi,jAre all greater than the cluster center separation distance Rc。
Counting and recording the number K of cluster centers in a cluster center pool to be selected and the geographic position of each cluster center corresponding to the micro cell base station, inputting the number K as a parameter into a traditional K-means to execute a K-means algorithm, and obtaining clustering results of all micro cell base stations in the ultra-dense network; referring to fig. 3, after the method of the present invention clusters the ultra-dense network including 50 microcell base stations, it is obvious from the clustering effect diagram that the clustering result of the scheme of the present invention is more uniform; wherein, 50 base stations are divided into 5 clusters in total, the base station number of each cluster is basically the same, and the clustering effect is ideal.
The invention relates to a base station clustering method based on density and minimum distance in an ultra-dense network, which comprises the steps of firstly calculating the distribution density and clustering density threshold of microcell base stations in the ultra-dense network, constructing an initial cluster center pool for the microcell base stations with the distribution density larger than the clustering density threshold, and calculating the distance between any two microcell base stations in the initial cluster center pool; then, calculating the weighted distribution density of each micro cell base station in the initial cluster center pool as the product of the actual distribution density and the minimum distance from the high-density base station, and arranging the base stations in the pool in a descending order according to the weighted distribution density to form a center pool to be selected; calculating the cluster center isolation distance of the pool to be selected; comparing the distance between any two micro cell base stations in the initial cluster center pool with the cluster center isolation distance, and removing the micro cell base stations smaller than the cluster center isolation distance from the cluster center pool to be selected; finally, counting and recording the number and the geographic position of the micro cell base stations in the central pool of the cluster to be selected, inputting the number and the geographic position as parameters of the traditional K-means, and executing a K-means algorithm to obtain a clustering result of the micro cell base stations in the ultra-dense network; compared with the prior art, the invention can adaptively set the number of clusters according to the position change of the base station in the network, thereby better realizing the clustering of the ultra-dense network; the initial cluster centers are jointly screened by calculating the distribution density and the minimum distance of the base station, so that the situation of falling into a local optimal solution can be avoided, and a clustering result with higher accuracy is obtained.
Although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments described in the foregoing detailed description, or equivalent changes may be made in some of the features of the embodiments described above. All equivalent structures made by using the contents of the specification and the attached drawings of the invention can be directly or indirectly applied to other related technical fields, and are also within the protection scope of the patent of the invention.
Claims (6)
1. A method for clustering base stations based on density and minimum distance in an ultra-dense network is characterized by comprising the following steps:
s1, recording the geographical positions of N micro cell base stations in the ultra-dense network, and calculating the Euclidean distance between any two micro cell base stations;
s2, calculating the distribution density and clustering density threshold of each micro cell base station in the ultra-dense network, comparing the distribution density of each micro cell base station with the clustering density threshold, and taking the micro cell base station with the distribution density larger than the clustering density threshold as an initial cluster center;
s3, forming an initial cluster center pool by all the initial cluster centers, and calculating the minimum value of the distance between each micro cell base station in the initial cluster center pool and the micro cell base station with higher distribution density;
s4, calculating the product of the distribution density of each micro cell base station in the initial cluster center pool and the minimum distance value between the micro cell base stations higher than the distribution density, and recording the product as weighted distribution density thetajAnd corresponding the micro cell base station to the weighted distribution density thetajThe cluster center pools to be selected are formed by descending order from large to small;
s5, calculating the cluster center separation distance and distributing the density theta according to the weightjThe cluster centers with smaller weighted distribution density value in the two cluster centers with the distance between every two cluster centers smaller than the cluster center isolation distance in the cluster center pool to be selected are sequentially removed from the cluster center pool to be selected;
s6, counting and recording the number K of cluster centers in the cluster center pool to be selected and the geographical position of each cluster center, inputting the number K and the geographical position of each cluster center as parameters into a traditional K-means algorithm, and executing the K-means algorithm to obtain clustering results of all micro cell base stations in the ultra-dense network.
2. The method for clustering base stations based on density and minimum distance in the ultra-dense network as claimed in claim 1, wherein in step S2, the distribution density is defined as:
the clustering density threshold is defined as:
3. The method for clustering base stations based on density and minimum distance in the ultra-dense network as claimed in claim 1, wherein in step S3, the larger the minimum value, the more uniform the distribution of the micro cell micro base stations in the center of the initial cluster is, and the minimum value of the distance between the micro cell base stations with the maximum distribution density in the center of the initial cluster is the distance between the micro cell base stations with the farthest distance;
and all the initial cluster centers in the initial cluster center pool are arranged in a descending order according to the distribution density of the micro cell base stations.
4. The method for clustering base stations based on density and minimum distance in ultra-dense network as claimed in claim 1, wherein in step S4, the weighted distribution density θjThe greater the description relates to the weighted distribution density θjThe larger the distribution density of the corresponding micro cell base station is, and the more the distribution density theta is corresponding to the weight distribution densityjThe farther the corresponding microcell base station is from the other microcell base stations.
5. The method for clustering base stations based on density and minimum distance in ultra-dense network as claimed in claim 2, wherein in step S5, the cluster center isolation distance is defined as:
6. The method for clustering base stations based on the density and the minimum distance in the super dense network as claimed in claim 5, wherein in step S5, the Euclidean distance d between any micro cell base station i and any micro cell base station j in the super dense network is determinedi,jIs separated from the cluster center by a distance RcComparison, if di,j<RcRemoving the micro cell base station j from the center pool of the cluster to be selected until the distance d between any two micro cell base stations in the center pool of the cluster to be selectedi,jAre all greater than the cluster center separation distance Rc。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910222536.0A CN109831794B (en) | 2019-03-22 | 2019-03-22 | Base station clustering method based on density and minimum distance in ultra-dense network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910222536.0A CN109831794B (en) | 2019-03-22 | 2019-03-22 | Base station clustering method based on density and minimum distance in ultra-dense network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109831794A CN109831794A (en) | 2019-05-31 |
| CN109831794B true CN109831794B (en) | 2022-03-11 |
Family
ID=66871074
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910222536.0A Active CN109831794B (en) | 2019-03-22 | 2019-03-22 | Base station clustering method based on density and minimum distance in ultra-dense network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109831794B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110505644B (en) * | 2019-09-26 | 2021-09-10 | 江南大学 | User task unloading and resource allocation joint optimization method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101267391A (en) * | 2008-03-27 | 2008-09-17 | 上海交通大学 | Wireless sensor network topology control method based on non-uniform sections |
| CN101959244A (en) * | 2010-09-29 | 2011-01-26 | 浙江工业大学 | Method for controlling hierarchical type route suitable for wireless sensor network |
| CN107659973A (en) * | 2017-08-23 | 2018-02-02 | 南京邮电大学 | Super-intensive network cluster dividing method based on density K means algorithms |
| CN108012275A (en) * | 2017-12-14 | 2018-05-08 | 重庆邮电大学 | Small base station user resource allocation methods based on dynamic clustering in super-intensive network |
| CN108632943A (en) * | 2018-03-30 | 2018-10-09 | 重庆邮电大学 | Cluster-dividing method based on small base station deployment density in 5G super-intensive networks |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090112533A1 (en) * | 2007-10-31 | 2009-04-30 | Caterpillar Inc. | Method for simplifying a mathematical model by clustering data |
-
2019
- 2019-03-22 CN CN201910222536.0A patent/CN109831794B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101267391A (en) * | 2008-03-27 | 2008-09-17 | 上海交通大学 | Wireless sensor network topology control method based on non-uniform sections |
| CN101959244A (en) * | 2010-09-29 | 2011-01-26 | 浙江工业大学 | Method for controlling hierarchical type route suitable for wireless sensor network |
| CN107659973A (en) * | 2017-08-23 | 2018-02-02 | 南京邮电大学 | Super-intensive network cluster dividing method based on density K means algorithms |
| CN108012275A (en) * | 2017-12-14 | 2018-05-08 | 重庆邮电大学 | Small base station user resource allocation methods based on dynamic clustering in super-intensive network |
| CN108632943A (en) * | 2018-03-30 | 2018-10-09 | 重庆邮电大学 | Cluster-dividing method based on small base station deployment density in 5G super-intensive networks |
Non-Patent Citations (1)
| Title |
|---|
| 超密集网络中基于分簇的无线网络资源分配技术研究;李文超;《万方数据库》;20181218;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109831794A (en) | 2019-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107333267A (en) | A kind of edge calculations method for 5G super-intensive networking scenes | |
| Chia et al. | Adaptive hybrid channel assignment in wireless mobile network via genetic algorithm | |
| Maksymyuk et al. | Fractal geometry based resource allocation for 5G heterogeneous networks | |
| CN107659973A (en) | Super-intensive network cluster dividing method based on density K means algorithms | |
| CN109863820B (en) | System and method for scalable radio network slicing | |
| EP3451731A1 (en) | Cell clustering method, spectrum overlapping multiplexing method based on method, and device thereof | |
| Cheng et al. | Joint user association and resource allocation in HetNets based on user mobility prediction | |
| CN106658526B (en) | Simulated annealing algorithm-based frequency spectrum allocation method in ultra-dense small cellular network | |
| Gang et al. | Inter-tenant resource sharing and power allocation in 5G virtual networks | |
| CN109831794B (en) | Base station clustering method based on density and minimum distance in ultra-dense network | |
| CN108366426A (en) | A kind of virtual map method based on simulated annealing under super-intensive environment | |
| CN105554771B (en) | Femto subscriber resource distribution method based on dense distribution | |
| CN110366187B (en) | Wireless network base station building site selection optimization method for balancing cost and signal intensity | |
| CN109600756B (en) | Physical cell identification and distribution method based on maximum degree priority dyeing algorithm | |
| CN108282888B (en) | D2D resource allocation method based on improved fuzzy clustering | |
| CN101808325B (en) | Method and device for allocating frequency spectrum | |
| CN108738028B (en) | Clustering method under ultra-dense networking | |
| Kashaf et al. | Self-organizing inter-cell interference coordination in 4G and beyond networks using genetic algorithms | |
| CN109673015B (en) | Tracking area planning method based on spectral clustering | |
| CN110380808B (en) | Micro-cell semi-clustering interference coordination method taking user equipment as center | |
| Xiaolan et al. | A Controller Placement Algorithm Based on Density Clustering in SDN | |
| CN102752860B (en) | Cooperation distributed resource allocation method in cordless communication network | |
| Pinagapany et al. | Solving channel allocation problem in cellular radio networks using genetic algorithm | |
| Xiao et al. | Minimum Separation Clustering Algorithm with High Separation Degree in Ultra-Dense Network | |
| CN108112047B (en) | User-centered semi-dynamic clustering method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |