Network topology generation method and device, electronic equipment and storage medium

Abstract

The method and device for generating a network topology, the electronic device and the storage medium provided by the present disclosure perform data recognition processing on target base station data obtained, to obtain new base station data; the new base station data is placed on a preset network topology for topology generation processing, to obtain a first network topology; the first network topology is input to a topology generation model for feature comparison processing, to obtain a first comparison result; the topology generation model contains topology layout features of the preset network topology; in a case where it is determined that any new base station data in the first network topology does not conform to the topology layout features, position adjustment processing is performed on the new base station data that does not conform to the topology layout features, to obtain a target network topology. Compared with related technologies, the embodiments of the present disclosure can locally adjust the layout of the network topology through the topology generation model, ensure that the layout of the network topology of other nodes is not affected, and generate a network topology that is correct in layout, beautiful, and better in user experience.

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, and in particular to a method and apparatus for generating network topology, an electronic device, and a storage medium. Background Technology

[0002] With the rapid expansion of operator networks and the number of wireless base stations, network management of wireless base stations has become increasingly important. As a crucial component of network management information, network topology provides essential foundational data for wireless network management and also provides vital support for intuitively displaying wireless network relationships. Wireless network topology typically requires viewing all subordinate wireless base stations of a certain type at the district / county or core network element granularity level, involving hundreds or even thousands of topology nodes. Existing network topology generation technologies mainly include: automatically laying out the network topology using force-oriented model-based layout algorithms, ring layout algorithms, and tree layout algorithms when users need to display the network topology of wireless base stations.

[0003] Because wireless base stations have many network elements and complex relationships, the network topology layout calculated directly using the above algorithm has problems such as overlapping network element nodes of wireless base stations and wireless base stations that are physically adjacent being placed far apart on the network topology. In this case, users need to manually drag and drop the network element nodes of wireless base stations on the network topology layout diagram to manually adjust them so that the network topology is aesthetically pleasing and scientific.

[0004] Meanwhile, the network elements of wireless base stations change frequently. For example, at the district / county level, new base stations are added or changed every one or two days, and the scope of each change is relatively small, generally not exceeding 1% of the total. If the network element nodes of a wireless base station change, recalculating the network topology layout using an algorithm based on the force-oriented model will cause the manually adjusted layout to be cleared. However, if the wireless topology layout is not recalculated, the changed network element nodes of the wireless base station cannot be well represented.

[0005] Therefore, while existing network topology generation technologies have solved the problem of network topology layout for wireless base stations to some extent, when there are a large number of wireless base stations, if a new network element node is added, the layout of all nodes in the network topology will be regenerated, resulting in layout errors, unsightly appearance, and poor user experience. Summary of the Invention

[0006] This disclosure provides a method, apparatus, electronic device, and storage medium for generating network topology. Its main objective is to address the shortcomings of existing network topology generation techniques. While these techniques have solved the problem of network topology layout for wireless base stations to some extent, when there are a large number of wireless base stations, adding new network element nodes requires regenerating the layout of all nodes in the network topology, leading to layout errors, unsightly appearance, and poor user experience.

[0007] According to a first aspect of this disclosure, a method for generating a network topology is provided, comprising:

[0008] The acquired target base station data is processed for data identification to obtain newly added base station data from the target base station data; wherein, the target base station data includes historical base station data and the newly added base station data;

[0009] The newly added base station data is placed on a preset network topology and topology generation is performed to obtain a first network topology; wherein, the preset network topology is the network topology corresponding to the historical base station data;

[0010] The first network topology is input into the topology generation model for feature comparison processing to obtain the first comparison result; wherein, the topology generation model contains the topological layout features of the preset network topology;

[0011] If any of the newly added base station data does not conform to the topology layout characteristics in the first network topology determined according to the first comparison result, the newly added base station data that does not conform to the topology layout characteristics is subjected to position adjustment processing to obtain the target network topology.

[0012] Optionally, after inputting the first network topology into the topology generation model for feature comparison processing to obtain the first comparison result, the method further includes:

[0013] If, in determining the first network topology based on the first comparison result, each of the newly added base station data conforms to the topology layout characteristics, the first network topology shall be used as the target network topology.

[0014] Optionally, the step of inputting the first network topology into the topology generation model for feature comparison processing to obtain the first comparison result includes:

[0015] In the first network topology, data interception processing is performed with each newly added base station data as the center to obtain the first image data corresponding to each newly added base station data.

[0016] The first image data is preprocessed using a preset processing algorithm to obtain the processed image data.

[0017] Each processed image data is compared with the topology layout features to obtain the first comparison result, wherein the first comparison result includes the comparison result corresponding to each newly added base station data.

[0018] Optionally, the step of performing feature comparison processing between each of the processed image data and the topological layout features to obtain the first comparison result includes:

[0019] Each processed image data is input into the first convolutional layer for data convolution processing to obtain the first feature vector corresponding to each processed image data; wherein, the topology generation model includes at least a convolutional layer, a pooling layer, a fully connected layer and an output layer;

[0020] The first feature vector is input into the first pooling layer for data pooling processing to obtain the pooled first feature vector, and the pooled first feature vector is activated according to the first preset activation function to obtain the first pooling vector.

[0021] The first pooling vector is input into the second convolutional layer for data convolution processing to obtain the second feature vector;

[0022] The second feature vector is input into the second pooling layer for data pooling processing to obtain the pooled second feature vector, and the pooled second feature vector is activated according to the first preset activation function to obtain the second pooling vector.

[0023] The second pooling vector is input into the third convolutional layer for data convolution processing to obtain the third feature vector, and the third feature vector is input into the fully connected layer for fully connected processing to obtain the target feature vector corresponding to each processed image data.

[0024] The target feature vector is input into the output layer and activated by a second preset activation function to obtain the similarity probability between each new base station data and the topology layout feature, and the first comparison result is determined based on the similarity probability.

[0025] Optionally, the step of performing location adjustment processing on the newly added base station data that does not conform to the topology layout characteristics to obtain the target network topology includes:

[0026] In the first network topology, the location adjustment processing is performed on the newly added base station data that does not conform to the topology layout characteristics to obtain the second network topology, and the second network topology is input into the topology generation model for feature comparison processing to obtain the second comparison result.

[0027] Based on the second comparison result, determine whether the data of each newly added base station in the second network topology conforms to the topology layout characteristics;

[0028] In determining the second network topology based on the second comparison result, if any of the newly added base station data does not conform to the topology layout characteristics, the newly added base station data that does not conform to the topology layout characteristics is subjected to position adjustment processing to obtain the third network topology;

[0029] The third network topology is input into the topology generation model for feature comparison processing to obtain a third comparison result. Based on the third comparison result, it is determined whether each of the newly added base station data in the third network topology conforms to the topology layout features.

[0030] The target network topology is obtained when it is determined that the data of each newly added base station conforms to the topology layout characteristics.

[0031] Optionally, the training method for the topology generation model includes:

[0032] In the preset network topology, data is extracted and processed with each historical base station data as the center to obtain the second image data corresponding to each historical base station data.

[0033] The second image data is preprocessed using a preset processing algorithm to obtain training image data.

[0034] The training image data is input into the model to be trained for model training to obtain the topology generation model.

[0035] Optionally, the step of inputting the training image data into the model to be trained for model training to obtain the topology generation model includes:

[0036] Each training image data is input into the first convolutional layer for data convolution processing to obtain the first training feature vector corresponding to each training image data;

[0037] The first training feature vector is input into the first pooling layer for data pooling processing to obtain the pooled first training feature vector. The pooled first training feature vector is then activated according to the first preset activation function to obtain the first training pooled vector.

[0038] The first training pooling vector is input into the second convolutional layer for data convolution processing to obtain the second training feature vector.

[0039] The second training feature vector is input into the second pooling layer for data pooling processing to obtain the pooled second training feature vector. The pooled second training feature vector is then activated according to the first preset activation function to obtain the second training pooled vector.

[0040] The second training pooling vector is input into the third convolutional layer for data convolution processing to obtain the third training feature vector, and the third training feature vector is input into the fully connected layer for fully connected processing to obtain the training target feature vector corresponding to each training image data.

[0041] The training target feature vector is input into the output layer and activated by the second preset activation function to obtain the training similarity probability between each new base station data and the topology layout feature;

[0042] Based on the similarity probabilities used in training, the model to be trained is optimized using a preset loss function to obtain the topology generation model.

[0043] According to a second aspect of this disclosure, a network topology generation apparatus is provided, comprising:

[0044] An identification unit is used to perform data identification processing on the acquired target base station data to obtain newly added base station data in the target base station data; wherein, the target base station data includes historical base station data and the newly added base station data;

[0045] A generation unit is used to place the newly added base station data on a preset network topology for topology generation processing to obtain a first network topology; wherein, the preset network topology is the network topology corresponding to the historical base station data;

[0046] The comparison unit is used to input the first network topology into the topology generation model for feature comparison processing to obtain a first comparison result; wherein, the topology generation model includes the topological layout features of the preset network topology;

[0047] An adjustment unit is configured to perform position adjustment processing on any newly added base station data that does not conform to the topology layout characteristics in the first network topology determined according to the first comparison result, so as to obtain a target network topology.

[0048] Optionally, the generation unit is further configured to, when each of the newly added base station data conforms to the topology layout characteristics in the first network topology determined according to the first comparison result, use the first network topology as the target network topology.

[0049] Optionally, the comparison unit includes:

[0050] The interception module is used to perform data interception processing in the first network topology, with each newly added base station data as the center, to obtain the first image data corresponding to each newly added base station data;

[0051] The processing module is used to preprocess the first image data using a preset processing algorithm to obtain processed image data.

[0052] The comparison module is used to perform feature comparison processing on each of the processed image data and the topology layout features to obtain the first comparison result, wherein the first comparison result includes the comparison result corresponding to each of the newly added base station data.

[0053] Optionally, the comparison module is further used for:

[0054] Each processed image data is input into the first convolutional layer for data convolution processing to obtain the first feature vector corresponding to each processed image data; wherein, the topology generation model includes at least a convolutional layer, a pooling layer, a fully connected layer and an output layer;

[0055] The first feature vector is input into the first pooling layer for data pooling processing to obtain the pooled first feature vector, and the pooled first feature vector is activated according to the first preset activation function to obtain the first pooling vector.

[0056] The first pooling vector is input into the second convolutional layer for data convolution processing to obtain the second feature vector;

[0057] The second feature vector is input into the second pooling layer for data pooling processing to obtain the pooled second feature vector, and the pooled second feature vector is activated according to the first preset activation function to obtain the second pooling vector.

[0058] The second pooling vector is input into the third convolutional layer for data convolution processing to obtain the third feature vector, and the third feature vector is input into the fully connected layer for fully connected processing to obtain the target feature vector corresponding to each processed image data.

[0059] The target feature vector is input into the output layer and activated by a second preset activation function to obtain the similarity probability between each new base station data and the topology layout feature, and the first comparison result is determined based on the similarity probability.

[0060] Optionally, the adjustment unit includes:

[0061] The adjustment module is used to perform position adjustment processing on the newly added base station data that does not conform to the topology layout characteristics in the first network topology to obtain a second network topology, and input the second network topology into the topology generation model for feature comparison processing to obtain a second comparison result;

[0062] The determining module is used to determine, based on the second comparison result, whether each of the newly added base station data in the second network topology conforms to the topology layout characteristics;

[0063] The adjustment module is further configured to, in the case that any of the newly added base station data in the second network topology determined according to the second comparison result does not conform to the topology layout characteristics, perform position adjustment processing on the newly added base station data that does not conform to the topology layout characteristics to obtain a third network topology;

[0064] The comparison module is used to input the third network topology into the topology generation model for feature comparison processing, obtain a third comparison result, and determine whether each of the newly added base station data in the third network topology conforms to the topology layout features based on the third comparison result.

[0065] The determining module is further configured to obtain the target network topology until it is determined that each of the newly added base station data conforms to the topology layout characteristics.

[0066] Optionally, the device further includes:

[0067] The interception unit is used to perform data interception processing in the preset network topology, with each historical base station data as the center, to obtain second image data corresponding to each historical base station data.

[0068] The processing unit is used to preprocess the second image data using a preset processing algorithm to obtain training image data;

[0069] The training unit is used to input the training image data into the model to be trained for model training processing to obtain the topology generation model.

[0070] Optionally, the training unit includes:

[0071] The convolution module is used to input each of the training image data into the first convolution layer for data convolution processing to obtain the first training feature vector corresponding to each of the training image data.

[0072] The pooling module is used to input the first training feature vector into the first pooling layer for data pooling processing to obtain the pooled first training feature vector, and to activate the pooled first training feature vector according to the first preset activation function to obtain the first training pooled vector.

[0073] The convolution module is also used to input the first training pooling vector into the second convolutional layer for data convolution processing to obtain the second training feature vector.

[0074] The pooling module is further configured to input the second training feature vector into the second pooling layer for data pooling processing to obtain the pooled second training feature vector, and to activate the pooled second training feature vector according to the first preset activation function to obtain the second training pooled vector.

[0075] The convolution module is further configured to input the second training pooling vector into the third convolution layer for data convolution processing to obtain the third training feature vector, and input the third training feature vector into the fully connected layer for fully connected processing to obtain the training target feature vector corresponding to each training image data.

[0076] The activation module is used to input the training target feature vector into the output layer, perform activation processing through a second preset activation function, and obtain the training similarity probability between each newly added base station data and the topology layout feature;

[0077] The optimization module is used to optimize the model to be trained based on the similarity probability used in training and through a preset loss function to obtain the topology generation model.

[0078] According to a third aspect of this disclosure, an electronic device is provided, comprising:

[0079] At least one processor; and

[0080] A memory communicatively connected to the at least one processor; wherein,

[0081] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect above.

[0082] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are configured to cause the computer to perform the method described in the first aspect above.

[0083] According to a fifth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method described in the first aspect above.

[0084] The network topology generation method, apparatus, electronic device, and storage medium disclosed herein perform data identification processing on acquired target base station data to obtain newly added base station data from the target base station data; wherein the target base station data includes historical base station data and the newly added base station data; the newly added base station data is placed on a preset network topology for topology generation processing to obtain a first network topology; wherein the preset network topology is the network topology corresponding to the historical base station data; the first network topology is input into a topology generation model for feature comparison processing to obtain a first comparison result; wherein the topology generation model includes the topology layout features of the preset network topology; if, based on the first comparison result, any newly added base station data in the first network topology does not conform to the topology layout features, the newly added base station data that does not conform to the topology layout features is subjected to position adjustment processing to obtain a target network topology. Compared with related technologies, the embodiments of this disclosure can intelligently identify the network element nodes of newly added wireless base stations. When generating the network topology, the layout of the network topology can be locally adjusted based only on the network element nodes of the newly added wireless base stations through the topology generation model. There is no need to regenerate the layout of all nodes in the network topology, ensuring that the layout of the network topology of other nodes is not affected, so as to generate a network topology with correct layout, good appearance and good user experience.

[0085] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0086] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0087] Figure 1 A schematic flowchart illustrating a method for generating a network topology according to an embodiment of this disclosure;

[0088] Figure 2 This is a schematic flowchart of a feature comparison process for a first network topology provided in an embodiment of the present disclosure;

[0089] Figure 3 This is a schematic diagram illustrating a target network topology generation process provided in an embodiment of the present disclosure.

[0090] Figure 4 This is a flowchart illustrating a training method for a topology generation model provided in an embodiment of the present disclosure.

[0091] Figure 5 This is a schematic diagram of the structure of a topology generation model provided in an embodiment of the present disclosure;

[0092] Figure 6 This is an overall flowchart of network topology generation provided in an embodiment of the present disclosure;

[0093] Figure 7 A schematic diagram of a network topology generation apparatus provided in an embodiment of this disclosure;

[0094] Figure 8 A schematic diagram of the structure of another network topology generation apparatus provided in an embodiment of this disclosure;

[0095] Figure 9 This is a schematic block diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0096] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0097] The following description, with reference to the accompanying drawings, outlines a method, apparatus, electronic device, and storage medium for generating network topology according to embodiments of the present disclosure.

[0098] Figure 1 This is a schematic flowchart illustrating a method for generating a network topology according to an embodiment of the present disclosure.

[0099] like Figure 1 As shown, the method includes the following steps:

[0100] Step 101: Perform data identification processing on the acquired target base station data to obtain newly added base station data in the target base station data; wherein, the target base station data includes historical base station data and the newly added base station data.

[0101] In this embodiment of the disclosure, the target base station data is the base station data obtained after updating the historical base station data. It can be understood that the historical base station data plus the newly added base station data is the target base station data. Therefore, the historical base station data can be all base station data in the target base station data except for the newly added base station data. If defined according to the time scale, the historical base station data can be the base station data at the moment before the target base station data, and the target base station data can be the base station data at the moment after the historical base station data.

[0102] Obtaining the target base station data only requires directly reading the currently stored base station data. When obtaining the new base station data, it can be obtained by comparing the target base station data with the historical base station data. Therefore, the step of performing data identification processing on the obtained target base station data to obtain the new base station data in the target base station data includes: performing data search processing in the historical base station data according to the specified base station data to obtain the data search result; wherein, the specified base station data is any base station data in the target base station data; if it is determined from the data search result that the specified base station data is found in the historical base station data, then the specified base station data is determined to be the historical base station data; if it is determined from the data search result that the specified base station data is not found in the historical base station data, then the specified base station data is determined to be the new base station data.

[0103] Specifically, regarding the acquisition of the newly added base station data, this disclosure also provides a programming algorithm for illustration, for example:

[0104] for each v in V{

[0105] / / Iterate through the nodes where the topology needs to be displayed

[0106] for each e in E{

[0107] / / Traverse the nodes of the original topology

[0108] if (v = e) {

[0109] exit;

[0110] / / If v finds any corresponding node in E, it means it is not a new node.

[0111] }

[0112] }

[0113] output v

[0114] / / If no corresponding node is found after traversing E, it means v is a new node.

[0115] }

[0116] Wherein, V is the designated base station data in the target base station data, and E is the historical base station data.

[0117] Step 102: Place the newly added base station data on a preset network topology and perform topology generation processing to obtain a first network topology; wherein, the preset network topology is the network topology corresponding to the historical base station data.

[0118] In this embodiment of the disclosure, the preset network topology is the network topology of the wireless base station currently in use, that is, the network topology corresponding to the historical base station data. When obtaining the preset network topology, it is only necessary to directly read the currently saved wireless network topology layout data.

[0119] Specifically, when placing the newly added base station data on the preset network topology, each newly added base station data is placed on the preset network topology, and each newly added base station data is placed in a different position on the preset network topology. That is, the newly added wireless nodes (the newly added base station data) are placed in different positions on the topology layout diagram (the preset network topology), thereby pre-generating the corresponding network topology (the first network topology).

[0120] It should be noted that when the newly added base station data is placed on the preset network topology, the position of each newly added base station data can be placed randomly, or it can be placed in a regular manner according to the topological layout characteristics of the preset network topology. Specifically, this disclosure embodiment does not impose any restrictions.

[0121] Step 103: Input the first network topology into the topology generation model for feature comparison processing to obtain the first comparison result; wherein, the topology generation model contains the topology layout features of the preset network topology.

[0122] In this embodiment of the disclosure, the purpose of performing feature comparison on the first network topology is to determine the feature conformity of the newly added base station data with the topology layout features when the data is located at different positions on the first network topology, and to determine the newly added base station data at the position with higher feature conformity and the newly added base station data at the position with lower feature conformity, which is the first comparison result.

[0123] The topology generation model is obtained by training the model through the preset network topology. Therefore, the topology generation model contains the topological layout features of the preset network topology. The structure of the topology generation model is a custom-selected model structure, such as a Convolutional Neural Network (CNN) model. Specifically, this embodiment does not limit the structure of the topology generation model.

[0124] Step 104: If any of the newly added base station data in the first network topology does not conform to the topology layout characteristics as determined by the first comparison result, the newly added base station data that does not conform to the topology layout characteristics is subjected to position adjustment processing to obtain the target network topology.

[0125] In this embodiment of the disclosure, if it is determined that any of the newly added base station data in the first network topology does not conform to the topology layout characteristics, that is, if it is determined that there are newly added base station data in the first network topology with low feature conformity, the newly added base station data in the low feature conformity location needs to be adjusted until each of the newly added base station data conforms to the topology layout characteristics, and the target network topology can be obtained.

[0126] When adjusting the location of the newly added base station data, the location can be adjusted by moving a preset distance of pixels each time. For example, the location can be moved 10 pixels at a time on the first network topology until the location that best matches the topology layout is found.

[0127] The network topology generation method disclosed herein involves performing data identification processing on acquired target base station data to obtain newly added base station data within the target base station data; wherein the target base station data includes historical base station data and the newly added base station data; placing the newly added base station data on a preset network topology for topology generation processing to obtain a first network topology; wherein the preset network topology is the network topology corresponding to the historical base station data; inputting the first network topology into a topology generation model for feature comparison processing to obtain a first comparison result; wherein the topology generation model includes topology layout features of the preset network topology; and if, based on the first comparison result, it is determined that any newly added base station data in the first network topology does not conform to the topology layout features, performing position adjustment processing on the newly added base station data that does not conform to the topology layout features to obtain a target network topology. Compared with related technologies, the embodiments of this disclosure can intelligently identify the network element nodes of newly added wireless base stations. When generating the network topology, the layout of the network topology can be locally adjusted based only on the network element nodes of the newly added wireless base stations through the topology generation model. There is no need to regenerate the layout of all nodes in the network topology, ensuring that the layout of the network topology of other nodes is not affected, so as to generate a network topology with correct layout, good appearance and good user experience.

[0128] In one possible implementation of this disclosure, when performing feature comparison processing on the first network topology, it refers to performing feature comparison processing on each newly added base station data in the first network topology to determine whether the location of the newly added base station data in the first network topology conforms to the topology layout features. Therefore, regarding the feature comparison processing of the first network topology, this disclosure provides a flowchart illustrating the feature comparison processing of the first network topology, as follows: Figure 2 As shown, it includes:

[0129] Step 201: In the first network topology, data interception processing is performed with each newly added base station data as the center to obtain the first image data corresponding to each newly added base station data.

[0130] In this embodiment of the disclosure, each newly added base station data is used as the center of the image data, and an image of a preset size is cropped in the first network topology to obtain the first image data corresponding to each newly added base station data. The image size of each first image data is a custom setting, such as 32*32, 64*64, etc. Specifically, this embodiment of the disclosure does not limit the image size of the first image data.

[0131] It should also be noted that the image size of the first image data is the same as the image size of the second image data. That is, the image size cropped centered on each of the newly added base station data needs to be consistent with the image size in the training image data used when training the topology generation model. For example, if the image size in the training image data is 32*32 when training the topology generation model, then when cropping data centered on each of the newly added base station data, an image of size 32*32 also needs to be cropped.

[0132] Step 202: Preprocess the first image data using a preset processing algorithm to obtain processed image data.

[0133] In this embodiment of the disclosure, the preset processing algorithm is a custom-selected algorithm, such as grayscale processing. Specifically, this embodiment of the disclosure does not limit the preset processing algorithm.

[0134] By preprocessing the first image data, on the one hand, the image lines are made clearer, and interference from other factors is reduced, making it easier for the topology generation model to extract the topology layout features of the wireless base station. On the other hand, the size of the image data is effectively reduced, the calculation time is reduced, and the calculation efficiency is improved.

[0135] It should also be noted that the preprocessing operations for the first image data are the same as those for the second image data. For example, when training the topology generation model, the second image data needs to be processed into grayscale to obtain training image data. Similarly, when performing feature comparison processing, the first image data also needs to be processed into grayscale to obtain processed image data for feature comparison.

[0136] Step 203: Perform feature comparison processing on each of the processed image data and the topology layout features to obtain the first comparison result, wherein the first comparison result includes the comparison result corresponding to each of the newly added base station data.

[0137] In this embodiment of the disclosure, when performing feature comparison, it is necessary to determine the comparison result of each newly added base station data. Therefore, it is necessary to perform feature comparison processing between each processed image data and the topology layout feature. Each processed image data can represent the position of each newly added base station data on the first network topology. The comparison result of each processed image data and the topology layout feature can determine whether each newly added base station data is in a position that conforms to the topology layout feature.

[0138] In one possible implementation of this disclosure, besides the possibility that any of the newly added base station data in the first network topology does not conform to the topology layout characteristics, there is also the possibility that every single newly added base station data in the first network topology conforms to the topology layout characteristics. When every single newly added base station data conforms to the topology layout characteristics, it indicates that every single newly added base station data is in a position that conforms to the topology layout characteristics. In this case, the first network topology does not need to be adjusted. Therefore, after obtaining the first comparison result, the following method can also be used, but is not limited to: if it is determined from the first comparison result that every single newly added base station data in the first network topology conforms to the topology layout characteristics, the first network topology is taken as the target network topology.

[0139] In one possible implementation of this embodiment, as a refinement of step 104 above, when any newly added base station data in the first network topology does not conform to the topology layout characteristics, it is necessary to adjust the position of the newly added base station data that does not conform to the topology layout characteristics. However, during position adjustment, there may be a situation where, after one adjustment, it still does not conform to the topology layout characteristics. Therefore, it is necessary to repeatedly adjust the position of the newly added base station data that does not conform to the topology layout characteristics until each of the newly added base station data conforms to the topology layout characteristics. Therefore, when performing position adjustment processing on the newly added base station data that does not conform to the topology layout characteristics, it can also be implemented in the following ways, but not limited to:

[0140] In the first network topology, the newly added base station data that does not conform to the topology layout characteristics are subjected to position adjustment processing to obtain a second network topology. The second network topology is then input into a topology generation model for feature comparison processing to obtain a second comparison result. Based on the second comparison result, it is determined whether each newly added base station data in the second network topology conforms to the topology layout characteristics. If, based on the second comparison result, any newly added base station data in the second network topology does not conform to the topology layout characteristics, the newly added base station data that does not conform to the topology layout characteristics is subjected to position adjustment processing to obtain a third network topology. The third network topology is then input into a topology generation model for feature comparison processing to obtain a third comparison result. Based on the third comparison result, it is determined whether each newly added base station data in the third network topology conforms to the topology layout characteristics. This process continues until it is determined that each newly added base station data conforms to the topology layout characteristics, at which point the target network topology is obtained.

[0141] In this embodiment of the disclosure, it is necessary to repeatedly adjust the position of newly added base station data that does not conform to the topology layout characteristics until each of the newly added base station data conforms to the topology layout characteristics. This can ensure that each newly added base station data in the target network topology is in a position that conforms to the topology layout characteristics, thereby achieving an ideal topology display effect and making the layout of the target network topology correct and aesthetically pleasing.

[0142] In summary, to facilitate understanding of the target network topology generation process, this disclosure provides a schematic diagram of the target network topology generation process, as shown in the above embodiments. Figure 3 As shown, the topology diagram is the preset network topology, the topology of the newly added wireless base station node is the first network topology, the 32*32 grayscale image is the processed image data, the wireless base station topology requirements are the topology layout features, and adjusting the coordinates of the newly added wireless base station node refers to adjusting the position of the newly added base station data that does not conform to the topology layout features.

[0143] In one possible implementation of this disclosure, as a refinement of step 203 above, when performing feature comparison processing between each processed image data and the topological layout features, the following methods can also be used, but are not limited to: inputting each processed image data into a first convolutional layer for data convolution processing to obtain a first feature vector corresponding to each processed image data; wherein, the topology generation model includes at least a convolutional layer, a pooling layer, a fully connected layer, and an output layer; inputting the first feature vector into a first pooling layer for data pooling processing to obtain a pooled first feature vector, and activating the pooled first feature vector according to a first preset activation function to obtain a first pooling vector; inputting the first pooling vector into a second convolutional layer. Data convolution is performed to obtain a second feature vector; the second feature vector is then input into a second pooling layer for data pooling to obtain a pooled second feature vector, and activated according to a first preset activation function to obtain a second pooled vector; the second pooled vector is then input into a third convolution layer for data convolution to obtain a third feature vector, and input into a fully connected layer for fully connected processing to obtain a target feature vector corresponding to each processed image data; the target feature vector is then input into an output layer and activated using a second preset activation function to obtain the similarity probability between each newly added base station data and the topology layout feature, and the first comparison result is determined based on the similarity probability.

[0144] In this embodiment of the disclosure, the topology generation model includes multiple convolutional layers, multiple pooling layers, a fully connected layer, and an output layer. The convolutional layers are used to filter data noise and extract key features. The pooling layers utilize the principle of local image correlation to subsample the image, which can reduce the amount of data processing while retaining useful information. The fully connected layer is used for classification or regression tasks based on the features extracted by the previous layers. The output layer is used to output the data.

[0145] The first preset activation function is a custom activation function, such as the reLU function, and the second activation function is also a custom activation function, such as the Softmax function. Specifically, this embodiment does not impose any restrictions on the first activation function and the second activation function.

[0146] In determining the first comparison result based on the similarity probability, the following methods may also be used, but are not limited to: if the similarity probability is greater than a preset probability threshold, determine that the new base station data with the similarity probability greater than the preset probability threshold conforms to the topology layout characteristics; if the similarity probability is less than or equal to the preset probability threshold, determine that the new base station data with the similarity probability less than or equal to the preset probability threshold does not conform to the topology layout characteristics.

[0147] In one possible implementation of this disclosure, the network topology needs to be generated using the topology generation model, which needs to be trained. Therefore, to obtain the topology generation model, this disclosure provides a flowchart illustrating a training method for the topology generation model, as shown below. Figure 4 As shown, it includes:

[0148] Step 401: In the preset network topology, data interception processing is performed with each historical base station data as the center to obtain the second image data corresponding to each historical base station data.

[0149] In this embodiment of the disclosure, each historical base station data is used as the center of the image data, and an image of a preset size is cropped from the preset network topology to obtain the second image data corresponding to each historical base station data. The image size of each second image data is a custom setting, such as 32*32, 64*64, etc. Specifically, this embodiment of the disclosure does not limit the image size of the second image data.

[0150] Step 402: Preprocess the second image data using a preset processing algorithm to obtain training image data.

[0151] In this embodiment of the disclosure, the preset processing algorithm is a custom-selected algorithm, such as grayscale processing. Specifically, this embodiment of the disclosure does not limit the preset processing algorithm.

[0152] By preprocessing the second image data, on the one hand, the image lines are made clearer, and interference from other factors is reduced, making it easier for the topology generation model to extract the topology layout features of the wireless base station. On the other hand, the size of the image data is effectively reduced, the calculation time is reduced, and the calculation efficiency is improved.

[0153] Step 403: Input the training image data into the model to be trained for model training processing to obtain the topology generation model.

[0154] In this embodiment of the disclosure, the model to be trained is an initial topology generation model. By training the model, the topology layout features of the current wireless base station nodes (the preset network topology) are trained, and the topology generation model can be obtained. The structure of the model to be trained is a custom-selected model structure, such as a Convolutional Neural Network (CNN) model, etc. Specifically, this embodiment of the disclosure does not limit the structure of the model to be trained.

[0155] The topology generation model is used to generate a network topology based on the new base station data, on the basis of the existing network topology (the preset network topology). The topology generation model is different for existing network topologies with different topology layout characteristics. In practical applications, different topology generation models can be trained according to existing network topologies with different topology layout characteristics.

[0156] In one possible implementation of this embodiment, as a refinement of step 403 above, when inputting the training image data into the model to be trained for model training, it can also be implemented in the following manner, but is not limited to: inputting each training image data into a first convolutional layer for data convolution processing to obtain a first training feature vector corresponding to each training image data; inputting the first training feature vector into a first pooling layer for data pooling processing to obtain a pooled first training feature vector, and activating the pooled first training feature vector according to a first preset activation function to obtain a first training pooled vector; inputting the first training pooled vector into a second convolutional layer for data convolution processing to obtain a second training feature vector; and inputting the second training feature vector into a second pooling layer. Data pooling is performed to obtain a pooled second training feature vector. This second training feature vector is then activated using a first preset activation function to obtain a second training pooled vector. This second training pooled vector is then input into a third convolutional layer for data convolution to obtain a third training feature vector. This third training feature vector is then input into a fully connected layer for fully connected processing to obtain a training target feature vector corresponding to each training image data. The training target feature vector is then input into the output layer and activated using a second preset activation function to obtain a training similarity probability between each newly added base station data and the topology layout features. Based on the training similarity probability, the model to be trained is optimized using a preset loss function to obtain the topology generation model.

[0157] In this embodiment of the disclosure, the topology generation model can classify the input data into two categories: those that conform to the topology layout features and those that do not conform to the topology layout features. In addition, during the training of the neural network, a loss function is needed to measure the performance of the topology generation model. The loss function is a custom-selected loss function, such as the cross-entropy loss function. Specifically, this embodiment of the disclosure does not limit the loss function.

[0158] Wherein, the first preset activation function is a custom activation function, such as the reLU function, and the second activation function is also a custom activation function, such as the Softmax function. Specifically, this embodiment does not impose any restrictions on the first activation function and the second activation function.

[0159] The topology generation model contains multiple layers. To intuitively understand the structure of the topology generation model, this disclosure provides a schematic diagram of the topology generation model, as shown below. Figure 5As shown, the system includes multiple convolutional layers, multiple pooling layers, a fully connected layer, and an output layer. The convolutional layers are used to filter data noise and extract key features. The pooling layers utilize the principle of local image correlation to subsample the image, which can reduce the amount of data processing while retaining useful information, thereby reducing the network training parameters and the overfitting degree of the model. The fully connected layer is used for classification or regression tasks based on the features extracted by the previous layers. The output layer is used to output the data.

[0160] It should be noted that the number of convolutional layers and pooling layers in the topology generation model is custom-configured. The number of convolutional layers and pooling layers is related to the size of the image data used when training the topology generation model. If the size of the second image data is large, the number of convolutional layers and pooling layers in the topology generation model will be large; if the size of the second image data is small, the number of convolutional layers and pooling layers in the topology generation model will be small.

[0161] For example, if the second image data is a 32*32 image, the topology generation model can contain 3 convolutional layers and 2 pooling layers, resulting in a 7-layer structure. If the second image data is a 64*64 image, the topology generation model can contain 4 convolutional layers and 3 pooling layers, resulting in a 9-layer structure. Specifically, the image size of the second image data and the number of layers in the topology generation model can be determined according to the actual situation, and this embodiment does not impose any limitations.

[0162] To facilitate understanding of the training process of the topology generation model, a 7-layer topology generation model is used as an example. In this embodiment, the specific method for constructing the CNN model (the topology generation model) is as follows: The CNN model includes 7 layers. The first layer is a first convolutional layer, used to filter noise and extract key features. Six 5x5 filters are used, with a stride s = 1 and padding = 0. The second layer is an average pooling layer (the first pooling layer), which utilizes the principle of local image correlation to subsample the image, reducing data processing while retaining useful information, and lowering network training parameters and model overfitting. In this embodiment, a 2x2 filter is used with a stride of s = 2 and padding = 0. The activation function of the second layer is reLU, and the value is taken as average pooling. The third layer is a second convolutional layer, using 16 5x5 filters with a stride of s = 1 and padding = 0. The fourth layer is a second pooling layer, using a 2x2 filter with a stride of s = 2 and padding = 0. The activation function of the fourth layer is still reLU, but the value is taken as max pooling. The fifth layer is a third convolutional layer, with 120 5x5 units, a stride of s = 1, and padding = 0. The sixth layer is a fully connected layer, which flattens the third convolutional layer of the fifth layer into a one-dimensional vector. The seventh layer is an output layer, which uses the Softmax function to output the probability of each category. In this embodiment, the categories are wireless base station node topology layout map conforming to or not conforming to the map. In addition, during the training of the neural network, a loss function is needed to measure the performance of the model. In this embodiment, cross-entropy loss is used to measure the difference between the predicted probability distribution and the true distribution.

[0163] Meanwhile, in combination with the above embodiments, when generating the network topology, it is necessary to pre-train the topology generation model based on the preset network topology corresponding to the historical base station data, then obtain the newly added base station data, and adjust the position of the newly added base station data based on the topology generation model to generate a target network topology that conforms to the topology layout characteristics.

[0164] For ease of understanding, this disclosure provides an overall flowchart for network topology generation, such as... Figure 6As shown, step s1 represents: obtaining the network topology corresponding to the historical base station data, i.e., the preset network topology; step s2 represents: in the preset network topology, data interception processing is performed with each historical base station data as the center to obtain the second image data corresponding to each historical base station data; the second image data is preprocessed using a preset processing algorithm to obtain training image data; step s3 represents: inputting the training image data into the model to be trained for model training processing to obtain the topology generation model; step s4 represents: performing data recognition processing on the acquired target base station data to obtain the newly added base station data in the target base station data; wherein, the target base station data includes historical base station data and the newly added base station data; step s5 represents: placing the newly added base station data in the preset network topology. A topology generation process is performed on the topology to obtain a first network topology; wherein, the preset network topology is the network topology corresponding to the historical base station data; the first network topology is input into the topology generation model for feature comparison processing to obtain a first comparison result; wherein, the topology generation model includes the topology layout features of the preset network topology; s6 indicates that: if, according to the first comparison result, any newly added base station data in the first network topology does not conform to the topology layout features, the newly added base station data that does not conform to the topology layout features is subjected to position adjustment processing to obtain a target network topology; if, according to the first comparison result, each newly added base station data in the first network topology conforms to the topology layout features, the first network topology is taken as the target network topology.

[0165] In summary, the embodiments disclosed herein can achieve the following effects:

[0166] 1. The embodiments of this disclosure can intelligently identify the network element nodes of newly added wireless base stations, and when generating the network topology, the layout of the network topology can be locally adjusted based only on the network element nodes of the newly added wireless base stations through the topology generation model, without having to regenerate the layout of all nodes in the network topology, ensuring that the layout of the network topology of other nodes is not affected, so as to generate a network topology with correct layout, aesthetics and good user experience.

[0167] 2. The embodiments of this disclosure utilize a topology generation model (CNN algorithm) to automatically select the optimal topology layout position for newly added wireless base station nodes. This allows for reference to the overall topology style, achieving an ideal topology display effect and significantly reducing the time required for manual intervention and adjustment.

[0168] 3. The wireless network topology generated in this embodiment only adjusts the topology layout locally for nodes that have changed, which effectively improves the pertinence of topology layout changes. The topology layout of nodes that have not changed is not affected, realizing that the wireless network topology generation has memory and growth. The overall topology adapts to the network adjustment, solving the problem that the overall topology layout needs to be regenerated after small changes in the network.

[0169] 4. The embodiments of this disclosure can intelligently identify newly added wireless base station nodes, and only need to adjust and adapt the topology layout of the newly added nodes, which greatly reduces the computing resources for generating wireless topology.

[0170] Corresponding to the network topology generation method described above, this invention also proposes a network topology generation apparatus. Since the apparatus embodiments of this invention correspond to the method embodiments described above, details not disclosed in the apparatus embodiments can be referred to in the method embodiments described above, and will not be repeated here.

[0171] Figure 7 This is a schematic diagram of the structure of a network topology generation apparatus provided in an embodiment of this disclosure, as shown below. Figure 7 As shown, it includes:

[0172] The identification unit 71 is used to perform data identification processing on the acquired target base station data to obtain newly added base station data in the target base station data; wherein, the target base station data includes historical base station data and the newly added base station data;

[0173] The generation unit 72 is used to place the newly added base station data on a preset network topology and perform topology generation processing to obtain a first network topology; wherein, the preset network topology is the network topology corresponding to the historical base station data;

[0174] The comparison unit 73 is used to input the first network topology into the topology generation model for feature comparison processing to obtain the first comparison result; wherein, the topology generation model includes the topological layout features of the preset network topology;

[0175] The adjustment unit 74 is used to perform position adjustment processing on the newly added base station data that does not conform to the topology layout characteristics in the first network topology determined according to the first comparison result, so as to obtain the target network topology.

[0176] The network topology apparatus provided in this disclosure performs data identification processing on acquired target base station data to obtain newly added base station data from the target base station data; wherein the target base station data includes historical base station data and the newly added base station data; the newly added base station data is placed on a preset network topology for topology generation processing to obtain a first network topology; wherein the preset network topology is the network topology corresponding to the historical base station data; the first network topology is input into a topology generation model for feature comparison processing to obtain a first comparison result; wherein the topology generation model includes the topology layout features of the preset network topology; if, based on the first comparison result, it is determined that any newly added base station data in the first network topology does not conform to the topology layout features, the newly added base station data that does not conform to the topology layout features is subjected to position adjustment processing to obtain a target network topology. Compared with related technologies, the embodiments of this disclosure can intelligently identify the network element nodes of newly added wireless base stations. When generating the network topology, the layout of the network topology can be locally adjusted based only on the network element nodes of the newly added wireless base stations through the topology generation model. There is no need to regenerate the layout of all nodes in the network topology, ensuring that the layout of the network topology of other nodes is not affected, so as to generate a network topology with correct layout, good appearance and good user experience.

[0177] Furthermore, in one possible implementation of this embodiment, the generation unit 72 is further configured to, when determining that each of the newly added base station data conforms to the topology layout characteristics in the first network topology based on the first comparison result, use the first network topology as the target network topology.

[0178] Furthermore, in one possible implementation of the embodiments of this disclosure, such as Figure 8 As shown, the comparison unit 73 includes:

[0179] The interception module 731 is used to perform data interception processing in the first network topology, with each newly added base station data as the center, to obtain the first image data corresponding to each newly added base station data;

[0180] Processing module 732 is used to preprocess the first image data using a preset processing algorithm to obtain processed image data;

[0181] The comparison module 733 is used to perform feature comparison processing on each of the processed image data and the topology layout features to obtain the first comparison result, wherein the first comparison result includes the comparison result corresponding to each of the newly added base station data.

[0182] Furthermore, in one possible implementation of this disclosure, the comparison module 733 is further configured to:

[0183] Each processed image data is input into the first convolutional layer for data convolution processing to obtain the first feature vector corresponding to each processed image data; wherein, the topology generation model includes at least a convolutional layer, a pooling layer, a fully connected layer and an output layer;

[0184] The first feature vector is input into the first pooling layer for data pooling processing to obtain the pooled first feature vector, and the pooled first feature vector is activated according to the first preset activation function to obtain the first pooling vector.

[0185] The first pooling vector is input into the second convolutional layer for data convolution processing to obtain the second feature vector;

[0186] The second feature vector is input into the second pooling layer for data pooling processing to obtain the pooled second feature vector, and the pooled second feature vector is activated according to the first preset activation function to obtain the second pooling vector.

[0187] The second pooling vector is input into the third convolutional layer for data convolution processing to obtain the third feature vector, and the third feature vector is input into the fully connected layer for fully connected processing to obtain the target feature vector corresponding to each processed image data.

[0188] The target feature vector is input into the output layer and activated by a second preset activation function to obtain the similarity probability between each new base station data and the topology layout feature, and the first comparison result is determined based on the similarity probability.

[0189] Furthermore, in one possible implementation of the embodiments of this disclosure, such as Figure 8 As shown, the adjustment unit 74 includes:

[0190] The adjustment module 741 is used to perform position adjustment processing on the newly added base station data that does not conform to the topology layout characteristics in the first network topology to obtain a second network topology, and input the second network topology into the topology generation model for feature comparison processing to obtain a second comparison result.

[0191] The determining module 742 is used to determine, based on the second comparison result, whether each of the newly added base station data in the second network topology conforms to the topology layout characteristics;

[0192] The adjustment module 741 is further configured to, in the case that any of the newly added base station data in the second network topology determined according to the second comparison result does not conform to the topology layout characteristics, perform position adjustment processing on the newly added base station data that does not conform to the topology layout characteristics to obtain a third network topology;

[0193] The comparison module 743 is used to input the third network topology into the topology generation model for feature comparison processing, obtain a third comparison result, and determine whether each of the newly added base station data in the third network topology conforms to the topology layout features based on the third comparison result.

[0194] The determining module 742 is further configured to obtain the target network topology until it is determined that each of the newly added base station data conforms to the topology layout characteristics.

[0195] Furthermore, in one possible implementation of the embodiments of this disclosure, such as Figure 8 As shown, the device further includes:

[0196] The interception unit 75 is used to perform data interception processing in the preset network topology, with each historical base station data as the center, to obtain second image data corresponding to each historical base station data.

[0197] Processing unit 76 is used to preprocess the second image data using a preset processing algorithm to obtain training image data;

[0198] Training unit 77 is used to input the training image data into the model to be trained for model training processing to obtain the topology generation model.

[0199] Furthermore, in one possible implementation of the embodiments of this disclosure, such as Figure 8 As shown, the training unit 77 includes:

[0200] The convolution module 771 is used to input each of the training image data into the first convolution layer for data convolution processing to obtain the first training feature vector corresponding to each of the training image data.

[0201] Pooling module 772 is used to input the first training feature vector into the first pooling layer for data pooling processing to obtain the pooled first training feature vector, and to activate the pooled first training feature vector according to the first preset activation function to obtain the first training pooled vector.

[0202] The convolution module 771 is further configured to input the first training pooling vector into the second convolution layer for data convolution processing to obtain the second training feature vector.

[0203] The pooling module 772 is further configured to input the second training feature vector into the second pooling layer for data pooling processing to obtain the pooled second training feature vector, and to activate the pooled second training feature vector according to the first preset activation function to obtain the second training pooled vector.

[0204] The convolution module 771 is further configured to input the second training pooling vector into the third convolution layer for data convolution processing to obtain the third training feature vector, and input the third training feature vector into the fully connected layer for fully connected processing to obtain the training target feature vector corresponding to each training image data.

[0205] The activation module 773 is used to input the training target feature vector into the output layer, perform activation processing through the second preset activation function, and obtain the training similarity probability between each newly added base station data and the topology layout feature;

[0206] The optimization module 774 is used to optimize the model to be trained based on the similarity probability used in training and through a preset loss function to obtain the topology generation model.

[0207] It should be noted that the foregoing explanation of the method embodiments also applies to the apparatus of the embodiments of this disclosure, and the principle is the same. Therefore, the embodiments of this disclosure are not limited thereto.

[0208] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0209] Figure 9 A schematic block diagram of an example electronic device 900 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0210] like Figure 9As shown, device 900 includes a computing unit 901, which can perform various appropriate actions and processes based on a computer program stored in ROM (Read-Only Memory) 902 or a computer program loaded from storage unit 908 into RAM (Random Access Memory) 903. RAM 903 can also store various programs and data required for the operation of device 900. The computing unit 901, ROM 902, and RAM 903 are interconnected via bus 904. I / O (Input / Output) interface 905 is also connected to bus 904.

[0211] Multiple components in device 900 are connected to I / O interface 905, including: input unit 906, such as keyboard, mouse, etc.; output unit 907, such as various types of monitors, speakers, etc.; storage unit 908, such as disk, optical disk, etc.; and communication unit 909, such as network card, modem, wireless transceiver, etc. Communication unit 909 allows device 900 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0212] The computing unit 901 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 901 include, but are not limited to, CPUs (Central Processing Units), GPUs (Graphics Processing Units), various special-purpose AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSPs (Digital Signal Processors), and any suitable processor, controller, microcontroller, etc. The computing unit 901 performs the various methods and processes described above, such as methods for generating network topologies. For example, in some embodiments, the network topology generation method can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 908. In some embodiments, part or all of the computer program can be loaded and / or installed on device 900 via ROM 902 and / or communication unit 909. When the computer program is loaded into RAM 903 and executed by the computing unit 901, one or more steps of the methods described above can be performed. Alternatively, in other embodiments, the computing unit 901 may be configured to perform the aforementioned network topology generation method by any other suitable means (e.g., by means of firmware).

[0213] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application-Specific Standard Products), SOCs (System-on-Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0214] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0215] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0216] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0217] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include LANs (Local Area Networks), WANs (Wide Area Networks), the Internet, and blockchain networks.

[0218] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service ecosystem, addressing the shortcomings of traditional physical hosts and VPS (Virtual Private Server, or simply "VPS") services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0219] It's important to note that artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, and big data processing. AI software technologies primarily include computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graph technologies.

[0220] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0221] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for generating network topology, characterized in that, include: The acquired target base station data is processed for data identification to obtain newly added base station data from the target base station data; wherein, the target base station data includes historical base station data and the newly added base station data; The newly added base station data is placed on a preset network topology and topology generation is performed to obtain a first network topology; wherein, the preset network topology is the network topology corresponding to the historical base station data; The first network topology is input into the topology generation model for feature comparison processing to obtain the first comparison result; wherein, the topology generation model contains the topological layout features of the preset network topology; In determining the first network topology based on the first comparison result, if any of the newly added base station data does not conform to the topology layout characteristics, the newly added base station data that does not conform to the topology layout characteristics is subjected to position adjustment processing to obtain the target network topology; In determining the first network topology based on the first comparison result, if each of the newly added base station data conforms to the topology layout characteristics, the first network topology shall be used as the target network topology. The step of inputting the first network topology into the topology generation model for feature comparison processing to obtain the first comparison result includes: In the first network topology, data interception processing is performed with each newly added base station data as the center to obtain the first image data corresponding to each newly added base station data. The first image data is preprocessed using a preset processing algorithm to obtain the processed image data. Each processed image data is compared with the topology layout features to obtain the first comparison result, wherein the first comparison result includes the comparison result corresponding to each newly added base station data.

2. The method according to claim 1, characterized in that, The step of performing feature comparison processing on each of the processed image data and the topological layout features to obtain the first comparison result includes: Each processed image data is input into the first convolutional layer for data convolution processing to obtain the first feature vector corresponding to each processed image data; wherein, the topology generation model includes at least a convolutional layer, a pooling layer, a fully connected layer and an output layer; The first feature vector is input into the first pooling layer for data pooling processing to obtain the pooled first feature vector, and the pooled first feature vector is activated according to the first preset activation function to obtain the first pooling vector. The first pooling vector is input into the second convolutional layer for data convolution processing to obtain the second feature vector; The second feature vector is input into the second pooling layer for data pooling processing to obtain the pooled second feature vector, and the pooled second feature vector is activated according to the first preset activation function to obtain the second pooling vector. The second pooling vector is input into the third convolutional layer for data convolution processing to obtain the third feature vector, and the third feature vector is input into the fully connected layer for fully connected processing to obtain the target feature vector corresponding to each processed image data. The target feature vector is input into the output layer and activated by a second preset activation function to obtain the similarity probability between each new base station data and the topology layout feature, and the first comparison result is determined based on the similarity probability.

3. The method according to claim 1, characterized in that, The step of performing location adjustment processing on the newly added base station data that does not conform to the topological layout characteristics to obtain the target network topology includes: In the first network topology, the location adjustment processing is performed on the newly added base station data that does not conform to the topology layout characteristics to obtain the second network topology, and the second network topology is input into the topology generation model for feature comparison processing to obtain the second comparison result. Based on the second comparison result, determine whether each of the newly added base station data in the second network topology conforms to the topology layout characteristics; In determining the second network topology based on the second comparison result, if any of the newly added base station data does not conform to the topology layout characteristics, the newly added base station data that does not conform to the topology layout characteristics is subjected to position adjustment processing to obtain the third network topology; The third network topology is input into the topology generation model for feature comparison processing to obtain the third comparison result. Based on the third comparison result, it is determined whether each of the newly added base station data in the third network topology conforms to the topology layout features. The target network topology is obtained when it is determined that the data of each newly added base station conforms to the topology layout characteristics.

4. The method according to claim 1, characterized in that, The training method for the topology generation model includes: In the preset network topology, data is extracted and processed with each historical base station data as the center to obtain the second image data corresponding to each historical base station data. The second image data is preprocessed using a preset processing algorithm to obtain training image data. The training image data is input into the model to be trained for model training to obtain the topology generation model.

5. The method according to claim 4, characterized in that, The step of inputting the training image data into the model to be trained for model training to obtain the topology generation model includes: Each training image data is input into the first convolutional layer for data convolution processing to obtain the first training feature vector corresponding to each training image data; The first training feature vector is input into the first pooling layer for data pooling processing to obtain the pooled first training feature vector. The pooled first training feature vector is then activated according to the first preset activation function to obtain the first training pooled vector. The first training pooling vector is input into the second convolutional layer for data convolution processing to obtain the second training feature vector. The second training feature vector is input into the second pooling layer for data pooling processing to obtain the pooled second training feature vector. The pooled second training feature vector is then activated according to the first preset activation function to obtain the second training pooled vector. The second training pooling vector is input into the third convolutional layer for data convolution processing to obtain the third training feature vector, and the third training feature vector is input into the fully connected layer for fully connected processing to obtain the training target feature vector corresponding to each training image data. The training target feature vector is input into the output layer and activated by the second preset activation function to obtain the training similarity probability between each new base station data and the topology layout feature; Based on the similarity probabilities used in training, the model to be trained is optimized using a preset loss function to obtain the topology generation model.

6. A network topology generation apparatus, characterized in that, include: An identification unit is used to perform data identification processing on the acquired target base station data to obtain newly added base station data in the target base station data; wherein, the target base station data includes historical base station data and the newly added base station data; A generation unit is used to place the newly added base station data on a preset network topology for topology generation processing to obtain a first network topology; wherein, the preset network topology is the network topology corresponding to the historical base station data; The comparison unit is used to input the first network topology into the topology generation model for feature comparison processing to obtain a first comparison result; wherein, the topology generation model includes the topological layout features of the preset network topology; An adjustment unit is configured to perform position adjustment processing on any newly added base station data that does not conform to the topology layout characteristics in the first network topology determined according to the first comparison result, so as to obtain a target network topology. The generation unit is further configured to, in the case that each of the newly added base station data conforms to the topology layout characteristics in the first network topology determined according to the first comparison result, use the first network topology as the target network topology; The comparison unit includes: The interception module is used to perform data interception processing in the first network topology, with each newly added base station data as the center, to obtain the first image data corresponding to each newly added base station data; The processing module is used to preprocess the first image data using a preset processing algorithm to obtain processed image data. The comparison module is used to perform feature comparison processing on each of the processed image data and the topology layout features to obtain the first comparison result, wherein the first comparison result includes the comparison result corresponding to each of the newly added base station data.

7. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5.

8. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-5.

9. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method according to any one of claims 1-5.

Images