A communication facility site selection method based on size constraint clustering, medium and terminal

Abstract

The application belongs to the field of communication technology and relates to a communication facility site selection method based on size-constrained clustering, a medium and a terminal, which comprises the following steps: S10, preprocessing before clustering; S20, size-constrained clustering; S30, AP site selection of a class: selecting a node as the position of an AP; S40, preprocessing of a class: connecting nodes capable of being connected with the AP and nodes capable of being interconnected with each other to form a large node, and connecting the large node set obtained through relaying; S50, limiting the search range of relays of a class; S60, relay site selection of a class: meeting the requirements of the maximum number of connectable nodes and the minimum average value of link budget attenuation; S70, isolated point processing: connecting isolated points with a class that has been planned through relaying, and attributing the isolated points to the class; S80, topology adjustment of a class; and S90, obtaining a multi-dimensional clustering result. The site selection method has high operation efficiency, and the clustering algorithm adopts a minimum cost flow model to obtain an optimal solution under the conditions of a fixed number of clusters and sizes.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of communication technology, and particularly relates to a communication facility site selection method based on size constraint clustering, a medium and a terminal. BACKGROUND

[0002] With the rapid development of China's communication industry, communication technology has been widely penetrated into all walks of life, and people's dependence on mobile communication has increased, and the requirements for communication speed, quality and safety have become higher and higher. China's communication operators have no exception to choose to increase the station capacity, so as to expand the network capacity and achieve the goal of seamless coverage. However, for the site selection of communication facilities in the planning area, it has been difficult to find a global optimal solution for a long time, and the current method can only get a feasible local solution, and it is impossible to confirm the goodness of the local solution. For example, in the prior art, the planning area is divided into several small areas according to the node position in the planning area, and then the position of the AP (wireless access point) in the small area is calculated according to the capacity of the AP and the node position. When the number of nodes in the small area exceeds the capacity of the AP, the small area is divided into two smaller areas, so that the number of nodes in the small area containing the AP is not greater than the capacity of the AP. According to this cycle, until the number of nodes in all small areas is not greater than the capacity of the AP.

[0003] The patent with publication number CN106507377B provides an optimal site selection method for relay stations in a communication network, comprising the following steps: step one, calculating the farthest node number ji of each node i without relay; for any node i∈V and i≠n, there is a unique value ji∈V, which represents the farthest node number that the signal can reach from the node; step two, calculating the lowest cost fi from each node and the corresponding relay station set Ui; in this step, the dynamic mathematical programming method is used to calculate the fi value of each node from the last node, that is, assuming that after establishing a relay station at the ith node, the lowest cost of all relay stations needed to be established in the middle to ensure that the signal from the ith node is transmitted to the target node B, including the relay station cost of node i; in addition, while calculating fi, the corresponding relay station set Ui corresponding to fi is also recorded; step three, determining the optimal relay station set; f1 is the lowest relay station construction cost, and U1 is the corresponding relay station set. The operation process in this patent is complex, the operation efficiency is low, and the result obtained cannot guarantee the optimal solution.

[0004] Therefore, how to provide a communication facility site selection method with high operation efficiency and guaranteed optimal solution is a problem to be solved by those skilled in the art. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application aims to provide a communication facility site selection method based on size-constrained clustering to solve the problem of low operation efficiency and difficulty in obtaining an optimal solution in the prior art, and further provides a communication facility site selection medium and terminal based on size-constrained clustering.

[0006] To solve the above technical problems, the present application adopts the following technical solutions:

[0007] In a first aspect, the present application provides a communication facility site selection method based on size-constrained clustering, comprising the following steps:

[0008] S10, preprocessing before clustering: taking a node as the center and R as the radius to draw a circle, the radius R can be adjusted, all nodes within the corresponding circle are regarded as a large node, and the process is repeated until all nodes are processed;

[0009] S20, size-constrained clustering: setting the number of clusters as N and the size as S, wherein the number of clusters N and the size S can be adjusted, N initial cluster centers are randomly selected, and a minimum cost flow model is used to obtain N new different classes;

[0010] S30, AP site selection of the class: selecting a node from the nodes of all new classes obtained in step S20 as the position of the AP;

[0011] S40, preprocessing of the class: connecting all nodes that can be connected with the AP and all nodes that can be interconnected between nodes, regarding the connected nodes as a large node, and connecting the obtained large node set through a relay;

[0012] S50, limit the search range of the relay of the class: finding the upper, lower, left and right boundaries of all new class nodes in step S20 to determine a rectangular region X, increasing a same value on the upper, lower, left and right boundaries of the rectangular region X, the value can be adjusted, expanding to a rectangular region Y, and the relays not in the rectangular region Y do not participate in the operation;

[0013] S60, relay site selection of the class: the position with the maximum number of connected nodes and the minimum average value of link budget decay is the installation position of the relay;

[0014] S70, isolated point processing: connecting the isolated points with a certain class planned by the relay, and regarding the isolated points as the class;

[0015] S80, topology adjustment of the class: re-planning the planned class, searching only the non-isolated points and relays in the class, starting from the AP, finding the nodes and relays that can be connected with the AP and setting them as the first level, finding the nodes and relays that can be connected with the first level and setting them as the second level, and so on, until all non-isolated points are connected;

[0016] S90, get multi-dimension clustering result.

[0017] Further, after the step S80, it further includes relay deduplication operation, and the specific steps are as follows:

[0018] S801, find out the relay set which needs to be deduplicated;

[0019] S802, select a relay from the relay set, the relay belongs to multiple classes, find out the class with the least hop number from the AP to the relay;

[0020] S803, find out the nodes and relays behind the relay through the deep-first search and attribute them to the class;

[0021] S804, return to the step S802, continue to select the next relay and execute the steps S802 to S803 until the relay set is traversed.

[0022] Further, before the step S10, it further includes setting of clustering parameters, and the clustering parameters include the number of clusters N, the cluster size S and the pre-processing radius R.

[0023] Further, after the step S20, it further includes fine-tuning of the clustering result, and if the K nodes near a node are mostly attributed to a class, the node is also attributed to the class through the K-neighbor algorithm.

[0024] Further, the multi-dimension clustering result in the step S90 includes the number and position of APs, relays and infrastructure, the hop number of nodes, isolated points and topology graph.

[0025] Further, in the step S30, the node which can be used as the position of the AP needs to satisfy that the sum of squares of distances of other nodes to the node is minimum.

[0026] Further, the step S70 includes the following specific steps:

[0027] S701, set the processing range radius R of a certain isolated point, and only search the available relays and class set in the range;

[0028] S702, find out the relay set which can be connected with the isolated point;

[0029] S703, if the relay set is an empty set, end the processing of the isolated point and start the processing of the next isolated point;

[0030] S704, disconnect the nodes and relays in the class set through the relay set, if one or more classes can be connected, select the best link budget, end the processing of the isolated point and start the processing of the next isolated point.

[0031] Further, in step S80, if there is still an unused relay in the class, the unused relay will be discarded.

[0032] In a second aspect, the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the method as described above.

[0033] In a third aspect, the present application also provides an electronic terminal, which comprises a processor and a memory, and the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to make the terminal execute the method as described above.

[0034] Compared with the prior art, the communication facility site selection method based on size-constrained clustering provided by the present application has at least the following beneficial effects:

[0035] The process of the present application is simple and efficient, and the node scale is reduced by preprocessing the node, so that the clustering can be directly completed by a common PC machine, the clustering algorithm adopts a minimum cost flow model, the operation efficiency is high, the multi-dimensional clustering result is obtained by automatically adjusting the clustering parameters and randomly selecting the clustering center, and then the global optimal solution can be obtained under the fixed clustering number and size conditions by repeatedly performing iteration operation on the clustering result. BRIEF DESCRIPTION OF DRAWINGS

[0036] In order to more clearly illustrate the scheme of the present application, the drawings needed in the following embodiment description will be simply introduced, and obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor on the basis of the drawings.

[0037] Figure 1 The overall flowchart of the communication facility site selection method based on size-constrained clustering provided by the embodiment of the present application is shown in the figure.

[0038] Figure 2 The schematic diagram of the preprocessing before clustering of the communication facility site selection method based on size-constrained clustering provided by the embodiment of the present application is shown in the figure.

[0039] Figure 3 The schematic diagram of the size-constrained clustering of the communication facility site selection method based on size-constrained clustering provided by the embodiment of the present application is shown in the figure.

[0040] Figure 4 The schematic diagram of the clustering result fine-tuning of the communication facility site selection method based on size-constrained clustering provided by the embodiment of the present application is shown in the figure.

[0041] Figure 5A schematic diagram of preprocessing of a class of a communication facility site selection method based on size-constrained clustering provided by an embodiment of the present application is shown in the figure.

[0042] Figure 6 A schematic diagram of limiting relay search range of a communication facility site selection method based on size-constrained clustering provided by an embodiment of the present application is shown in the figure.

[0043] Figure 7 A schematic diagram of relay site selection of a communication facility site selection method based on size-constrained clustering provided by an embodiment of the present application is shown in the figure.

[0044] Figure 8 A schematic diagram of isolated point processing of a communication facility site selection method based on size-constrained clustering provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0045] In order to facilitate the understanding of the present application, the present application will be described in more detail below with reference to the relevant drawings. The preferred embodiments of the present application are shown in the drawings. However, the present application can be realized in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present application more thorough and comprehensive.

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terms used in the specification of the present application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application.

[0047] The present application provides a communication facility site selection method based on size-constrained clustering, which is applied to the site selection work of wireless access points AP, relays and other communication facilities in a planning area. The communication facility site selection method based on size-constrained clustering comprises the following steps:

[0048] S10, preprocessing before clustering: taking a certain node as the center and R as the radius to draw a circle, the radius R can be adjusted, and all nodes within the corresponding circle are regarded as a large node, and the process is repeated until all nodes are processed;

[0049] S20, size-constrained clustering: setting the number of clusters as N and the size as S, wherein the number of clusters N and the size S can be adjusted, N initial cluster centers are randomly selected, and a minimum cost flow model is used to finally obtain N new different classes;

[0050] S30, AP site selection of a class: selecting a node from the nodes of all new classes obtained in step S20 as the position of the AP;

[0051] S40. Class preprocessing: Connect all nodes that can be connected to the AP and nodes that can be interconnected with each other, and treat the connected nodes as a large node. Connect the resulting large node set through relays.

[0052] S50. Limit the search range of relays in the class: using the node coordinates of all new classes in step S20, find the upper, lower, left and right boundaries of all new classes, determine a rectangular area X, and add the same value to the upper, lower, left and right sides of the rectangular area X. This value can be adjusted and expanded to the rectangular area Y. Relays not in the rectangular area Y do not participate in the calculation.

[0053] S60. Relay location selection: The location with the most connectable nodes and the lowest average link budget attenuation is the relay installation location;

[0054] S70, Outlier handling: Use relays to connect outliers to a pre-defined class and assign the outlier to that class;

[0055] S80. Class topology adjustment: Re-plan the planned class, only search for non-isolated points and relays in the class. Starting from AP, first find the nodes and relays that can be connected to AP and set them as the first level. Then find the nodes and relays that can be connected to AP from the first level and set them as the second level. And so on, until all non-isolated points are connected.

[0056] S90. Obtain multi-dimensional clustering results.

[0057] The location selection method in this invention has high computational efficiency, and the clustering algorithm adopts the minimum cost flow model to obtain an optimal solution under the condition of fixed number and size of clusters.

[0058] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0059] This invention provides a method for site selection of communication facilities based on size-constrained clustering, applicable to the site selection of wireless access points (APs), relays, and other communication facilities within a planned area. Figure 1 As shown, the communication facility location method based on size-constrained clustering includes the following steps:

[0060] S10. Preprocessing before clustering: Draw a circle with a certain node as the center and R as the radius. The radius R can be adjusted. Treat all nodes inside the circle as a large node, and so on, until all nodes have been processed.

[0061] Specifically, clustering is performed only on large nodes, which reduces the node size, thereby improving computational efficiency and saving resource consumption. Figure 2 As shown, we will eventually get three large nodes: A, B, and C. Clustering is only performed on A, B, and C.

[0062] S20, size-constrained clustering: set the number of clusters N and the size S, where the number of clusters N and the size S are adjustable, randomly select N initial cluster centers, and finally obtain N clusters using the minimum cost flow model;

[0063] Through the minimum cost flow model, regardless of the initial cluster center, a result close to the global optimum will be obtained, specifically, as shown in Figure 3 , let N = 3, S = 3, and the initial cluster centers be A, B, and C. Finally, three clusters 1, 2, and 3 will be obtained.

[0064] Further, after clustering, the clustering result is fine-tuned using the K- nearest neighbor algorithm. If the majority of the K nodes near a node belong to a certain class, then the node also belongs to this class. Through fine-tuning, the points on the class boundary can be better attributed.

[0065] Specifically, as shown in Figure 4 , take K = 3, and the three points near A all belong to class 2, so A will be attributed to class 2.

[0066] S30, AP selection for class: select a node from the nodes of the class as the location of the AP, which satisfies the minimum sum of squared distances from other nodes to this node;

[0067] This strategy can ensure that the AP is located at the center of all nodes, and the overall effect is optimal.

[0068] S40, class preprocessing: connect all nodes that can be connected to the AP and nodes that can be interconnected between nodes, and consider the connected nodes as a large node. The large node set obtained is the one that needs to be connected by a relay. If there is only one large node, no relay is needed;

[0069] Through class preprocessing, the size of the operation node can be reduced, and the operation efficiency can be improved, as shown in Figure 5 , after class preprocessing, there are only three large nodes A, B, and C, and only A, B, and C need to be connected by a relay.

[0070] S50, limit the search range of the class relay: find the upper, lower, left, and right boundaries of the class nodes to determine a rectangular region X. The upper, lower, left, and right of X are increased by the same value, which can be adjusted, and expanded to a rectangular region Y. Relays not in Y will not participate in the operation;

[0071] By limiting the range, global search is avoided, and the operation efficiency is greatly improved, as shown in Figure 6 , the boundary rectangle of the class is ABCD, which is expanded to the rectangle EFGH. Relay 6 and relay 7 are not in EFGH, and will not participate in the operation.

[0072] S60, relay location selection of the class: the location with the most connectable nodes and the smallest average link budget attenuation is the relay installation location;

[0073] Using the greedy strategy, each step is optimal, as shown in Figure 7 A, B, C, D, E are large nodes that need to be connected by relays, and the specific steps are as follows:

[0074] S601, find all relays 1, 2, 3 that can be connected to A;

[0075] S602, use relays 1, 2, 3 to connect BCDE (DE is out of range, and will not be connected by link budget), relay 1 can only connect node B, relays 2 and 3 can connect BC, but the average link budget of relay 2 connecting BC is greater than that of relay 3, so finally relay 3 will be selected;

[0076] S603, judge whether there are isolated points: if not, end; if yes, proceed to step S604;

[0077] S604, find relays 4, 5 that can be connected to relay 3, at this time the available relays are relays 1, 2, 4, 5, and the remaining nodes are DE;

[0078] S605, jump to step S602 (replace relays 1, 2, 3 with relays 1, 2, 4, 5, and replace BCDE with DE), and finally relay 5 will be selected to connect the remaining DE.

[0079] S70, isolated point processing: try to connect the isolated point to a certain class that has been planned by using a relay, and classify the isolated point into this class;

[0080] Specifically, the steps are as follows:

[0081] S701, set the processing range radius R of a certain isolated point A (R can be adjusted), and only search for available relays and class sets (a set of planned classes) within this range, which can improve processing efficiency;

[0082] S702, find the relay set S1 that can be connected to A;

[0083] S703, if the relay set S1 is empty, end the processing of the isolated point and start the processing of the next isolated point;

[0084] S704, connect the nodes in the class set and the relays through the relay set S1, if one or more classes are found to be connected, select the one with the best link budget, end the processing of the isolated point and start the processing of the next isolated point; if not found, find the relay set S2 that can be connected through the relay set S1;

[0085] S705, S1=S2, return to step S703.

[0086] Through processing, isolated points can be reduced, and the final solution is more optimal, and specifically, as shown in the figure, the processing of isolated point A, relay 1, class 5 will not search, and finally A, B and C three isolated points are no longer isolated. Figure 8

[0087] S80, topology adjustment of the class: strict breadth first, re-planning the planned class, searching only the non-isolated points and relays in the class, starting from the AP, first finding the nodes and relays that can be connected to the AP, setting them as the first level, then finding the nodes and relays that can be connected to the first level, setting them as the second level, and so on until all non-isolated points are connected, at this time, there may be relays in the class that are not used, and these relays are no longer needed;

[0088] Through adjustment, the number of relays is further optimized, and topology analysis can be performed accordingly.

[0089] Further, after the topology adjustment of the class, the relay deduplication strategy is performed: because different classes of operations may all place relays on the same infrastructure (such as high-voltage poles and low-voltage poles for installing relays), the specific steps are as follows:

[0090] S801, find the relay set S1 that needs to be deduplicated;

[0091] S802, select a relay A from S1, the relay A belongs to multiple classes, and find the class X with the least number of hops from the AP to the relay A;

[0092] S803, through depth first search, find all the nodes and relays behind the relay A, and classify them into class X;

[0093] S804, return to step S802, continue with the next relay B until the relay set S1 is traversed.

[0094] Through deduplication, the nodes are reclassified, the planning can be further optimized, and the number of relays can be reduced.

[0095] S90, obtain multi-dimensional clustering results: including the number and position of APs, relays and infrastructures; the number of hops, isolated points and topology graph of nodes.

[0096] Through multi-dimensional results, the cost and operation performance can be considered comprehensively.

[0097] The embodiment of the application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement any one of the methods in the embodiment.

[0098] ​The embodiment of the present application further provides an electronic terminal, comprising a processor and a memory; the memory is used for storing a computer program, and the processor is used for executing the computer program stored in the memory, so that the terminal executes any method in the embodiment.

[0099] The computer readable storage medium in the embodiment can be understood by those skilled in the art that all or part of the steps of the above-mentioned method embodiments can be completed by a computer program related hardware. The foregoing computer program can be stored in a computer readable storage medium. The program executes the steps of the above-mentioned method embodiments when executed; and the foregoing storage medium includes ROM, RAM, magnetic disc or optical disc and various storage program codes.

[0100] The electronic terminal provided by the embodiment comprises a processor, a memory, a transceiver and a communication interface, the memory and the communication interface are connected with the processor and the transceiver and complete communication between each other, the memory is used for storing a computer program, the communication interface is used for communication, and the processor and the transceiver are used for running the computer program, so that the electronic terminal executes each step of the method.

[0101] The communication facility site selection method, medium and terminal based on size constraint clustering described in the above embodiment take probability theory as a theoretical basis, take efficient operation and low resource consumption as a basis, and have simple flow and efficient process. By preprocessing the nodes, the node scale is reduced, so that the clustering can be directly completed by a common PC machine. The clustering algorithm adopts a minimum cost flow model, has high operation efficiency, obtains a multi-dimensional clustering result by automatically adjusting the clustering parameters and randomly selecting a clustering center, and then repeatedly iteratively operates the clustering result, so that the global optimal solution can be obtained under the condition of fixed clustering number and size.

[0102] Obviously, the above-described embodiments are only the preferred embodiments of the present application, but not all the embodiments. The preferred embodiments of the present application are shown in the drawings, but do not limit the patent scope of the present application. The present application can be realized in many different forms, and the embodiments are provided to make the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art can modify the technical solutions recorded in the foregoing embodiments, or equivalently replace part of the technical features. Any equivalent structure made by using the contents of the present application specification and drawings, directly or indirectly applied to other related technical fields, is also within the patent protection scope of the present application.

Claims

1. A method for site selection of communication facilities based on size-constrained clustering, characterized in that, The method comprises the following steps: S10, pre-processing before clustering: taking a node as the center and R as the radius to draw a circle, all nodes in the corresponding circle are regarded as a large node, and the process is repeated until all nodes are processed; S20, size-constrained clustering: setting the number of clusters as N and the size as S, randomly selecting N initial cluster centers, and obtaining N new different classes by using a minimum cost flow model; S30, AP site selection of the class: selecting a node from all nodes of the new class obtained in step S20 as the position of the AP; In step S30, the node selected as the position of the AP needs to satisfy that the sum of squares of distances from other nodes to the node is minimum; S40, pre-processing of the class: connecting nodes that can be connected with the AP and nodes that can be interconnected between nodes, regarding the connected nodes as a large node, and connecting the obtained large node set through a relay; S50, search range limitation of the relay of the class: finding the upper, lower, left and right boundaries of all nodes of the new class in step S20, determining a rectangular region X, and expanding the rectangular region X to a rectangular region Y by adding a same value to the upper, lower, left and right boundaries of the rectangular region X, so that the relay not in the rectangular region Y does not participate in the operation; S60, relay site selection of the class: the position with the maximum number of connectable nodes and the minimum average value of link budget attenuation is the installation position of the relay; S70, isolated point processing: connecting the isolated point with a certain class planned by the relay, and regarding the isolated point as the class; S80, topology adjustment of the class: re-planning the planned class, searching only non-isolated points and relays in the class, starting from the AP, finding out nodes and relays that can be connected with the AP and setting them as the first level, finding out nodes and relays that can be connected with the first level and setting them as the second level, and iteratively performing the operation until the non-isolated points are connected together; S90, obtaining a multi-dimensional clustering result; The multi-dimensional clustering result in step S90 comprises the number and position of APs, relays and infrastructure, the hop number of nodes, isolated points and a topology graph.

2. The method of claim 1, wherein, After step S80, a relay de-duplication operation is further included, and the specific steps are as follows: S801, finding out a relay set that needs to be de-duplicated; S802, selecting a relay from the relay set, the relay belongs to multiple classes, and finding out a class with the minimum hop number from the AP to the relay; S803, finding out nodes and relays behind the relay by depth-first search and regarding them as the class; S804, returning to step S802, continuing to select the next relay and performing steps S802 to S803 until the relay set is traversed.

3. The method of claim 2, wherein, Before step S10, a setting of a clustering parameter is further included, and the clustering parameter comprises a cluster number N, a cluster size S and a pre-processing radius R.

4. The method of claim 3, wherein, After step S20, a fine adjustment of a clustering result is further included, and the K-nearest neighbor algorithm is used to fine adjust the clustering result, if the majority of K nodes near a node belong to a certain class, the node also belongs to the class.

5. The method of claim 1, wherein, The specific steps of step S70 are as follows: S701, setting a processing range radius R of a certain isolated point, and searching only available relays and class sets in the range. S702, find out the relay set which can be connected with the isolated point; S703, if the relay set is empty set, end the processing of the isolated point and start the processing of the next isolated point; S704, through the relay set, disconnect the nodes and relays in the class set, if one or more classes are found to be connected, select the best link budget, end the processing of the isolated point and start the processing of the next isolated point.

6. The method of claim 1, wherein, In step S80, if there are unused relays in the class, the unused relays will be discarded.

7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, and the computer program is executed by the processor to implement the method in any one of claims 1 to 6.

8. An electronic terminal, characterized in that Comprise: A processor and a memory; The memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory, so that the terminal executes the method in any one of claims 1 to 6.

Images