Base station site selection method, electronic device, storage medium and computer program product

Abstract

The invention discloses a base station site selection method, an electronic device, a storage medium and a computer program product, and relates to the technical field of wireless communication, in the application, a target area is divided into candidate grid areas, so that the granularity of site selection is refined. Meanwhile, according to the corresponding grid fusion features of the candidate grid areas, the area site selection fitness of the grids is determined, the area site selection fitness is used as a basis for subsequently determining a target grid area, and finally, the position of a target base station is determined from the target grid area to complete site selection. It can be understood that compared with a related technical scheme for determining the position of the base station according to individual dimensions, the method can generate the site selection basis of the base station based on the fusion data (namely the fusion result of the grids in all aspects) of the grids, greatly enriches the reference basis of the site selection of the base station, guarantees the comprehensiveness of the site selection basis of the base station, and improves the site selection accuracy of the base station. Therefore, the optimal base station position can be selected as much as possible.

Description

Technical Field

[0001] The present application relates to the field of wireless communication technology, and in particular to a base station site selection method, electronic equipment, storage medium and computer program product. Background Art

[0002] In mobile communication networks, although existing base station optimization positioning technologies provide certain foundations and functions, they have obvious shortcomings. The data dimensions for site selection in related technologies are limited, and site selection is usually based on data from individual dimensions. For example, in related technologies, the distance between the base station and the sampling point is inferred based on the timing advance (TA) of the sampling point of the minimization of drive tests (MDT). However, the base station signal received at the actual sampling point is likely not line-of-sight propagation, and the base station distance inferred based on TA will be far greater than the actual distance. Therefore, if the location of the base station is determined based on the individual dimensions of TA, the accuracy of the base station site selection will be low, and it is difficult to ensure that the optimal base station location can be determined.

[0003] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0004] The main purpose of this application is to provide a base station site selection method, electronic device, storage medium and computer program product, aiming to solve the technical problem that the accuracy of base station site selection in related technologies is not high and it is difficult to ensure that the optimal base station location can be determined.

[0005] To achieve the above object, the present application proposes a base station site selection method, which comprises the following steps:

[0006] Divide the target area into candidate grid areas;

[0007] For any one of the candidate grid regions, determining the regional site selection adaptability of the candidate grid region based on the grid fusion features of the candidate grid region;

[0008] After traversing the candidate grid areas, based on the regional site selection adaptability of the candidate grid areas, a target grid area is screened out from the candidate grid areas, and the target base station position is determined according to the target grid area.

[0009] Optionally, the grid fusion feature includes a first fusion feature and a second fusion feature, and the step of determining the regional site selection adaptability of the candidate grid area based on the grid fusion feature of the candidate grid area includes:

[0010] Extracting multi-dimensional information features of the candidate grid area;

[0011] Fusing the multi-dimensional information features to obtain a first fused feature of the candidate grid region;

[0012] Extracting user behavior information and network load characteristics corresponding to the candidate grid area;

[0013] Performing feature fusion based on the user behavior information and the network load feature to obtain a second fusion feature;

[0014] Extracting terrain features and building features of the candidate grid area;

[0015] The regional site selection adaptability of the candidate grid area is determined based on the first fusion feature, the second fusion feature, the terrain feature, and the building feature.

[0016] Optionally, the step of determining the regional site selection adaptability of the candidate grid area based on the first fusion feature, the second fusion feature, the terrain feature, and the building feature includes:

[0017] Determining the network coverage quality of the candidate grid area by using the first fusion feature, the second fusion feature, the terrain feature, and the building feature;

[0018] The regional site selection adaptability of the candidate grid area is obtained by performing a weighted calculation based on the network coverage quality and the construction cost of building a base station in the candidate grid area.

[0019] Optionally, the step of determining the network coverage quality of the candidate grid area by using the first fusion feature, the second fusion feature, the terrain feature, and the building feature includes:

[0020] Normalization processing is performed on the first fusion feature, the second fusion feature, the terrain feature, and the building feature respectively to obtain normalized results, and the normalized results are accumulated to obtain the network coverage quality.

[0021] Optionally, the multidimensional information features include measurement data of a signaling network management system, measurement data of a geographic information system, weather data, and traffic flow data, and the step of fusing the multidimensional information features to obtain a first fused feature of the candidate grid area includes:

[0022] Performing weighted calculation on the measurement data of the signaling network management system, the measurement data of the geographic information system, the weather data, and the traffic flow data to obtain the first fusion feature; or

[0023] The measurement data of the signaling network management system, the measurement data of the geographic information system, the weather data and the traffic flow data are input into a preset fusion model to obtain the first fusion feature.

[0024] Optionally, the step of determining the target base station location according to the target grid area includes:

[0025] generating each candidate base station location based on the target grid area;

[0026] Determining the base station site selection fitness of each candidate base station location, and selecting a pending base station location from the candidate base station locations according to the base station site selection fitness;

[0027] Based on the pending base station position, new candidate base station positions are derived, and based on the new candidate base station positions, the step of determining the base station site fitness of each candidate base station position is returned to execute until a preset number of iterations is reached, and the final pending base station position is used as the target base station position.

[0028] Optionally, after the step of determining the location of the target base station according to the target grid area, the method further includes:

[0029] Determining signal quality information of the base station after the base station is built at the target base station location;

[0030] The operating parameters of the base station built at the target base station location are optimized based on the quality information.

[0031] In addition, to achieve the above objectives, the present application also proposes an electronic device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the base station site selection method as described above.

[0032] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the base station site selection method described above are implemented.

[0033] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the base station site selection method described above are implemented.

[0034] One or more technical solutions proposed in this application have at least the following technical effects:

[0035] In an embodiment of the present application, the target area is divided into candidate grid areas; for any one of the candidate grid areas, the regional site selection fitness of the candidate grid area is determined based on the grid fusion features of the candidate grid area; after traversing the candidate grid areas, the target grid area is screened out from the candidate grid areas based on the regional site selection fitness of the candidate grid areas, and the target base station location is determined based on the target grid area. It should be noted that in the present application, the target area will be divided into candidate grid areas to refine the granularity of the site selection. At the same time, the regional site selection fitness of the grid is determined based on the corresponding grid fusion features of the candidate grid area, and the regional site selection fitness will be used as the basis for the subsequent determination of the target grid area, and finally the target base station location is determined from the target grid area to complete the site selection. It can be understood that compared with the related technical solutions that determine the location of base stations based on individual dimensions, the present application will generate the basis for base station site selection based on the fusion data of the grid (that is, through the fusion results of various aspects of the grid), which greatly enriches the reference basis for base station site selection and ensures the comprehensiveness of the basis for base station site selection. Therefore, the present application can select the optimal base station location as much as possible. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0038] Figure 1 This is a flow chart of the first embodiment of the base station site selection method of the present application;

[0039] Figure 2 This is a flow chart of the second embodiment of the base station site selection method of the present application;

[0040] Figure 3 This is a schematic diagram of the flow chart of the third embodiment of the base station site selection method of the present application;

[0041] Figure 4 This is a flow chart of the fourth embodiment of the base station site selection method of the present application;

[0042] Figure 5 A schematic diagram of the device structure of the hardware operating environment involved in the base station site selection method in the embodiment of the present application.

[0043] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0044] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0045] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0046] In mobile communication networks, although existing base station optimization positioning technologies provide certain foundations and functions, they have obvious shortcomings. The data dimensions for site selection in the existing technology are limited, and the site selection is usually based on data from individual dimensions. For example, in the relevant technology, the distance between the base station and the sampling point is inferred based on the timing advance (TA) of the sampling point of the minimization of drive tests (MDT). However, the base station signal received by the actual sampling point is likely not line-of-sight propagation. The base station distance inferred based on TA will be far greater than the actual distance. Therefore, if the location of the base station is determined based on the individual dimensions of TA, the accuracy of the base station site selection will be low, and it is difficult to ensure that the optimal base station location can be determined.

[0047] The main solution of the embodiment of the present application is: dividing the target area into candidate grid areas; for any one of the candidate grid areas, determining the regional site selection fitness of the candidate grid area based on the grid fusion characteristics of the candidate grid area; after traversing the candidate grid areas, based on the regional site selection fitness of the candidate grid areas, screening out the target grid area from the candidate grid areas, and determining the target base station location according to the target grid area.

[0048] In this application, the target area will be divided into candidate grid areas to refine the granularity of the site selection. At the same time, the regional site selection fitness of the grid is determined based on the corresponding grid fusion characteristics of the candidate grid area, and the regional site selection fitness will be used as the basis for the subsequent determination of the target grid area, and finally the target base station location is determined from the target grid area to complete the site selection. It can be understood that compared with the relevant technical solutions for determining the location of the base station according to individual dimensions, this application will generate the location selection basis of the base station based on the fusion data of the grid (that is, through the fusion results of various aspects of the grid), which greatly enriches the reference basis for the site selection of the base station and ensures the comprehensiveness of the site selection basis of the base station. Therefore, this application can select the optimal base station location as much as possible.

[0049] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a cloud platform, a computer, a mobile phone, etc., or an electronic device that can realize the above functions.

[0050] It should be noted that the process of selecting a site for a base station in the present application involves data collection, and accordingly, also involves a process of data cleaning, so the process of data cleaning is first described.

[0051] Step 1: Efficient MR data cleaning algorithm, where MR represents RSRP and RSRQ, which represent ReferenceSignalReceived Power, reference signal received power; ReferenceSignalReceivedQuality, reference signal received quality.

[0052] The system (such as a system that applies the base station site selection method, which can also be the executor of the base station site selection method of the present application) collects network performance data in real time, including MR data, device status information, and network traffic data, by connecting to multiple data source interfaces.

[0053] Data collection is not limited to traditional base station data, but also includes data from cells, user equipment and core networks, providing data support for comprehensive network status. The detailed fields are shown in Table 1 below:

[0054]

[0055]

[0056] It is worth noting that in actual applications, the types and quantities of data collected may be more or less than the above examples, so they will not be repeated here.

[0057] Step 1.1 Missing value cleaning.

[0058] For missing values ​​in the data, choose to fill or delete. For example, if the RSRP or RSRQ data of a certain sampling point is missing, you can fill it with the average value of the records of adjacent sampling points, or directly delete these records.

[0059] Step 1.2: Outlier cleaning.

[0060] Use IQR (Interquartile Range) or standard deviation method to identify outliers. For example, calculate the IQR of RSRP and RSRQ in real time and remove data points that fall outside 1.5*IQR. Use the following formula:

[0061] [IQR = Q3 - Q1]

[0062] Where Q1 is the first quartile, i.e., the data point at the 25% position of the data set; Q3 is the third quartile, i.e., the data point at the 75% position of the data set.

[0063] [Upper=Q3+1.5×IQR, Lower=Q1-1.5×IQR]

[0064] In the formula, Q1 is the first quartile, that is, the data point at the 25% position of the data set; Q3 is the third quartile, that is, the data point at the 75% position of the data set; Upper represents data higher than 1.5*IQR; Lower represents data lower than 1.5*IQR.

[0065] Step 1.3 Data normalization.

[0066] Fields such as RSRP and RSRQ are normalized to ensure that the data is analyzed on the same scale. Common normalization methods include Z-score normalization.

[0067] Statistical methods are used to identify outliers, and the Zscore of each data point is calculated in real time to assess its degree of deviation from the mean. The real-time calculation formula for the Zscore of a data point is:

[0068]

[0069] Among them, (X) is the value of a single data point in the collected data set, (μ) is the average value of all data in the data set, and (σ) is the standard deviation. Usually, when the absolute value of the z value calculated by the above formula is greater than the preset threshold (such as 2), the corresponding data point is regarded as an outlier and removed from the data set.

[0070] The cleaned data is standardized to unify the data format and scale for subsequent data analysis and processing. Data standardization methods include MinMax Scaling, which scales the data to between 0 and 1. The formula is as follows:

[0071]

[0072] Where (X min ) and (X max ) are the minimum and maximum values ​​in the data set, respectively. This standardization method helps to eliminate the impact of data of different magnitudes and ensure the consistency of data processing.

[0073] Step 1.4: Cleaning in a unified format.

[0074] Ensure that all fields (such as Cell ID) are in the same format, for example, convert all Cell IDs to string format and remove any illegal characters. Verify the continuity and logic of the timestamp field and correct any obviously wrong time stamps, such as future timestamps or timestamps that are out of order.

[0075] The entire data cleaning process described above can be executed through automated scripts in the system, including data loading, outlier detection, data denoising, standardization, etc. The scripts can be run regularly to ensure that the data is continuously updated and prepared, providing support and preparation for subsequent data fusion and analysis.

[0076] Based on the above introduction, the present application embodiment provides a base station site selection method, referring to Figure 1 , which is a flow chart of the first embodiment of the base station site selection method of the present application.

[0077] In this embodiment, the base station site selection method includes steps S10 to S30:

[0078] Step S10, dividing the target area into candidate grid areas;

[0079] It should be noted that the target area refers to the area where a base station needs to be built, which is usually determined based on actual business needs. In order to facilitate site selection, in this embodiment, the target area will be divided to obtain candidate grid areas.

[0080] For example, in practical applications, a larger geographic area can be subdivided into smaller, more manageable units, i.e., grids, by using pre-set division rules, thereby optimizing the data integration and analysis process. For example, the geographic boundaries of the entire target area can be defined first, including the minimum and maximum longitudes (London min ), (Lon max )) and Latitude ((Lat min ), (Lat max )). These boundaries determine the total area that will be rasterized.

[0081] The grid size is preset according to the needs of network optimization and the level of detail of geographic information. This size determines the geographic range of each grid, and each grid is set to represent a specific area on the ground, such as 50x50 meters.

[0082] For any point in the area, the coordinates of the grid where it is located are calculated in real time based on its latitude and longitude positions.

[0083]

[0084] In the formula, Longitude and Latitude represent the longitude and latitude of the point respectively, GridSizelon and GridSize lat are the sizes of the grid in longitude and latitude respectively. x and G y Rounding is performed to ensure that each point is uniquely mapped to a raster.

[0085] Each grid is based on its coordinates G x and G y is assigned a unique identifier (ID), which can be simply represented by (ID = G y ×N+G x ), where N is the total number of grids in the longitude direction of the rasterized area. This unique ID is essential for subsequent data management and query.

[0086] All collected MR data are mapped to corresponding grids according to their geographical location information. The mapped grids are also the candidate grid areas mentioned above.

[0087] Step S20, for any one of the candidate grid regions, determining the regional site selection adaptability of the candidate grid region based on the grid fusion features of the candidate grid region;

[0088] It should be noted that, since the process of determining the regional site selection fitness according to each candidate grid region is substantially the same, in this embodiment, one of the candidate grid regions will be taken as an example for description.

[0089] Exemplarily, for any candidate grid area among the candidate grid areas (the candidate grid area may also be referred to as a grid in the subsequent description), the grid fusion features of the candidate grid area are extracted, and the regional site selection fitness of the candidate grid area is calculated based on the extracted grid fusion features. Among them, the above-mentioned grid fusion features can be obtained by fusion based on the data mapped by the candidate grid area (that is, the data obtained by data cleaning). For example, it can be RSRP, RSRQ, RSSI, etc. of the candidate grid area. The fusion process can be to normalize the data of various aspects of the candidate grid area and assign weights for weighted fusion. Then calculate the regional site selection fitness based on the grid fusion features. The regional site selection fitness can represent the score of the base station site selection in the candidate grid area. Correspondingly, the higher the regional site selection fitness, the higher the probability of the base station site selection in the candidate grid area. In addition, for the calculation process of the regional site selection fitness, the technicians can refer to the above rules and set it according to actual needs, so it will not be repeated here.

[0090] It can be understood that compared with the related technical solutions, which determine the location of the base station according to individual dimensions, the present application will generate the basis for the site selection of the base station (i.e. the above-mentioned regional site selection adaptability) based on the fusion data of the grid (the fusion result of the data of each dimension of the grid), which greatly enriches the reference basis for the site selection of the base station and improves the comprehensiveness of the site selection of the base station, so as to select the best base station location as much as possible.

[0091] Step S30, after traversing the candidate grid areas, based on the regional site selection adaptability of the candidate grid areas, a target grid area is screened out from the candidate grid areas, and a target base station position is determined according to the target grid area.

[0092] It should be noted that after traversing each candidate grid area, the regional location fitness of each candidate grid area can be obtained. Then, according to the regional location fitness of each candidate grid area, the target grid area is screened out from each candidate grid area. For example, the candidate grid area with the highest regional location fitness is screened out as the target grid area. If multiple target grid areas need to be screened out, the N candidate grid areas with the highest regional location fitness can also be used as the target grid areas. Finally, the target base station location is determined from the target grid area. For example, the center position of the target grid area is used as the target position, or the position in the surrounding area of ​​the center position of the target grid area that can be used to build a base station is used as the target position, etc.

[0093] In this embodiment, the target area is divided into candidate grid areas; for any one of the candidate grid areas, the regional site selection fitness of the candidate grid area is determined based on the grid fusion features of the candidate grid area; after traversing the candidate grid areas, the target grid area is screened out from the candidate grid areas based on the regional site selection fitness of the candidate grid areas, and the target base station location is determined based on the target grid area. It should be noted that in this application, the target area will be divided into candidate grid areas to refine the granularity of the site selection. At the same time, the regional site selection fitness of the grid is determined based on the corresponding grid fusion features of the candidate grid area, and the regional site selection fitness will be used as the basis for the subsequent determination of the target grid area, and finally the target base station location is determined from the target grid area to complete the site selection. It can be understood that compared with the related technical solutions that determine the location of base stations based on individual dimensions, the present application will generate the basis for base station site selection based on the fusion data of the grid (that is, through the fusion results of various aspects of the grid), which greatly enriches the reference basis for base station site selection and ensures the comprehensiveness of the basis for base station site selection. Therefore, the present application can select the optimal base station location as much as possible.

[0094] Reference Figure 2, is a flow chart of a second embodiment of the base station site selection method of the present application, which is proposed based on the first embodiment of the base station site selection method of the present application. In this embodiment, the same or similar contents as the above embodiments can be referred to the above introduction, and will not be repeated later. The grid fusion feature includes a first fusion feature and a second fusion feature, and the step of determining the regional site selection adaptability of the candidate grid area based on the grid fusion feature of the candidate grid area includes steps S21 to S26:

[0095] Step S21, extracting multi-dimensional information features of the candidate grid area;

[0096] Step S22, fusing the multi-dimensional information features to obtain a first fused feature of the candidate grid region;

[0097] Step S23, extracting user behavior information and network load characteristics corresponding to the candidate grid area;

[0098] Step S24, performing feature fusion based on the user behavior information and the network load feature to obtain a second fusion feature;

[0099] Step S25, extracting terrain features and building features of the candidate grid area;

[0100] Step S26: determining the regional site selection adaptability of the candidate grid area based on the first fusion feature, the second fusion feature, the terrain feature, and the building feature.

[0101] It should be noted that the above-mentioned grid fusion features may include a first fusion feature and a second fusion feature, and the first fusion feature and the second fusion feature are respectively obtained by fusing different grid features mapped to the candidate grid area, thereby organically fusing different grid features.

[0102] Exemplarily, firstly, the multi-dimensional information features of the candidate grid area are extracted, wherein the multi-dimensional information features may include measurement data of the signaling network management system (such as RSRP and RSRQ), measurement data of the geographic information system (GIS, Geographic Information System) (topography, building location), weather data and traffic flow data, etc. Then, the multi-dimensional information features are fused to obtain the first fused feature of the candidate grid area. Similarly, weights can be assigned to the features of each dimension first, and then the features of each dimension can be normalized and weighted summed to obtain the first fused feature.

[0103] Extract the user behavior information and network load characteristics corresponding to the candidate grid area, and then fuse the user behavior information and network load characteristics to obtain the second fusion feature. The user behavior information may include user movement speed, user stay time, and access frequency. The specific process can be as follows (i.e., dynamic user distribution modeling):

[0104] Regarding the user's moving speed, the speed can be calculated in real time based on the timestamps and geographic coordinates of two consecutive locations of the user in the candidate grid area. The speed V can be calculated in real time using the following formula:

[0105]

[0106] Where d is the distance between two consecutive locations (the distance between two points on the earth can be calculated in real time using the Haversine formula), and Δt is the time difference between the two recorded locations.

[0107] The user's stay time, that is, the total time the user stays in the candidate grid area, can be calculated in real time by counting the timestamps of the user entering and leaving the grid area. For each grid, the time difference between the user's entry and exit is calculated in real time, and all the stay times in the grid are accumulated to obtain the user's stay time in the corresponding candidate grid area.

[0108] Regarding user access frequency. The access frequency of users in each grid can be counted, and the grid with the highest number of accesses can be identified. For example, density estimation (Kernel Density Estimation, KDE) can be used to establish a distribution model of users in a specific area. This is a non-parametric method for estimating the probability density function:

[0109] KDE is a method for estimating the probability density function of a random variable. It smoothly summarizes the location of the data points and does not depend on the distribution form of the data. The density estimation formula corresponding to KDE is as follows:

[0110]

[0111] Where K is the kernel function (such as Gaussian kernel), X i is the sample point, h is the user traffic, controls the smoothness of the estimate, and n is the number of samples.

[0112] KDE is used to estimate user distribution and can help identify densely populated areas or hotspots of user activity. Selecting the right user traffic is key, because too little traffic will lead to overfitting (that is, too rugged estimates), while too much traffic may lead to overly smooth estimates that fail to capture the true structure of the data. Through this feature engineering and statistical modeling approach, network operators can more accurately understand and predict user behavior and distribution, making network resource allocation more efficient.

[0113] The extraction of network load characteristics (i.e. establishing base station network load characteristics) is as follows:

[0114] In the base station site selection process of this application, predicting the base station network load is a key link, which involves the analysis of a large amount of historical and real-time data. This process uses time features, user activity features, geographic location features and derived features to establish an accurate prediction feature. The following is the specific process:

[0115] 1. Establish time characteristics.

[0116] For example, a day can be divided into multiple periods such as morning peak, noon, and evening peak, and network usage patterns and load conditions can be analyzed based on users' daily activity patterns.

[0117] Divide the 24 hours of a day into several key periods, usually related to user activity patterns. For example: morning peak: 7:00-9:00AM, noon: 12:00-2:00PM, evening peak: 5:00-7:00PM, late night period: 12:00-5:00AM (low traffic period)

[0118] Temporal feature extraction: Add temporal features to each data record, including the hour and a binary indicator of whether it is a peak hour (1 for yes, 0 for no).

[0119] Historical load data statistics: Real-time calculation of historical average load and variability indicators for each period. Among them, historical average load indicators mainly include real-time calculation of average signal quality (such as RSRP and RSRQ), average number of users and data throughput in the same period of the past few weeks. Variability indicators mainly include related indicators that significantly deviate from historical average load indicators, which are recorded as obvious outliers.

[0120] Geographic location encoding: Extract geographic location-related features, such as the type of area (commercial area, residential area, etc.), based on the grid ID where the user is located.

[0121] 2. Establish user activity characteristics.

[0122] Analyze session duration, frequency, and data usage to categorize users (high, medium, and low activity) and use these categorizations to predict network demand.

[0123] User activity classification: First, users are classified based on their session duration, session frequency, and data usage. The specific steps are as follows:

[0124] 2.1, Session duration: Calculate the average duration of each user session in real time.

[0125] 2.2, Session frequency: Real-time calculation of the number of user sessions within a specific time window (such as per day or per week).

[0126] 2.3, Data usage: Statistics of the user’s total data consumption within a certain period of time.

[0127] 2.4, Population density index: Calculate the population density of each grid, the formula is:

[0128]

[0129] 2.5, Regional Activity Intensity: Evaluate the activity intensity of each grid based on the distribution of commercial, educational, residential and other activities in the region. This can be achieved by weighting activity points, where each activity type has a different weight.

[0130]

[0131] Among them, (w i ) is the weight of activity type (i), (a i ) is the number of activities of this type.

[0132] 3. Establish geographic location features.

[0133] Using grid ID to identify the specific location of the user and analyzing it based on the type of area (commercial area, residential area, etc.) to which the location belongs is an effective method to predict the data traffic demand of a specific area in different time periods.

[0134] 3.1, Grid ID Identification: Each user’s location is identified by a grid ID, which is a method of breaking down a geographic location into smaller blocks, with each grid representing a specific geographic area.

[0135] 3.2, Area Type Determination: According to the grid ID and the data characteristics in the corresponding area, the area is divided into several different types of areas, including commercial areas, residential areas, industrial areas, etc. This classification is based on the main use of the area and the characteristics of human flow.

[0136] 4. Create derived features.

[0137] The moving average load, load change trend, feature fusion and processing, and feature enhancement methods are combined to improve the prediction accuracy and reliability.

[0138] 4.1 Establish moving average load characteristics.

[0139] Moving average load is achieved by calculating the average network load over the past 7 or 30 days in real time. This method helps predict the average load conditions that a base station may face in the future by smoothing daily fluctuations. The formula used is as follows:

[0140]

[0141] Where n is the number of days (7 or 30 days), daily load i is the load on day i.

[0142] 4.2 Establish load change trend characteristics.

[0143] In this embodiment, it involves an in-depth study of the load changes in the same historical time period to identify network load fluctuations caused by specific time periods or events (such as holidays, special public events). The following is a detailed description of the technical solution:

[0144] 4.2.1 Time period analysis:

[0145] Group data by timestamp to distinguish between working and non-working days, as well as specific holidays and public events.

[0146] Aggregate the data in each time period and calculate the average network load in real time, including total data traffic and the maximum number of user connections.

[0147] 4.2.2 Trend analysis and pattern recognition:

[0148] A time series analysis technique, seasonally decomposed time series forecasting model (STL), is used to analyze the trend and seasonal pattern of network load.

[0149] Use machine learning techniques such as cluster analysis to identify load patterns in different time periods, especially abnormal changes during major events or holidays. The algorithm formula is:

[0150] [Load t =β0+β1 Holiday t +β2 Event t +∈ t ]

[0151] Among them, Load t represents the network load at time t, Holiday tand Event t are binary indicator variables representing holidays and specific events, β0, β1, β2 are model parameters, ∈ t is the error term.

[0152] Correspondingly, in this embodiment, the above network load characteristics may include time characteristics, user activity characteristics, geographic location characteristics, and derived characteristics.

[0153] Exemplarily, the user behavior information and the network load feature are fused to obtain a second fused feature (i.e., feature fusion and processing) as follows:

[0154] Through the above feature engineering steps, MR data can be effectively converted into features with high predictive value, which reflect the network usage pattern and load of a region (grid or candidate grid area) in different time periods. Such analysis helps network operators optimize resource allocation and network management, so as to better cope with the changes in data demand at different times and regions.

[0155] 1. Calculate average flow in real time.

[0156] The average flow of the grid in different time windows is calculated in real time, thus providing data support for network operators to optimize network resources and services.

[0157] Different time windows can be defined based on the needs of network operators, such as hourly or daily. This allows operators to adjust the length of the time window based on operational strategies and user behavior patterns.

[0158] Aggregate the data in each grid ID and calculate the average network load in the specified time window. The specific aggregation formula is:

[0159]

[0160] The data volume is the total data usage of all users in the time window.

[0161] 2. Analyze the traffic differences between business districts on weekdays and weekends.

[0162] Define a time window (e.g. daily) and aggregate the data by weekdays and weekends. Calculate the following metrics for each time window: average number of users, total data usage, average session duration.

[0163] The aggregation formula is as follows:

[0164]

[0165] Among them, day represents the date type (weekday or weekend), t represents the time window, and n is the number of users in the time window. i，t is the amount of data of the ith user in time window t.

[0166] The statistical test ANOVA was used to analyze the significance of the difference in traffic data between weekdays and weekends in the commercial district. These tests helped to confirm whether the difference in traffic between the two day types was statistically significant.

[0167] The specific algorithm is as follows:

[0168]

[0169] in, and are the average traffic on weekdays and weekends, S p is the combined standard deviation of the two samples, n is the number of samples in each group, and t is the difference in traffic data in the commercial area on weekdays and weekends.

[0170] Based on the analysis results, specific suggestions for optimizing network resources are made. For example, if it is found that the traffic on weekdays is significantly higher than on weekends, it is recommended to increase the network bandwidth on weekdays or deploy temporary mobile base stations during high-demand periods.

[0171] 3. Integrate demographic data with regional activity patterns.

[0172] Combine the population density index and the regional activity intensity to create a fusion feature that reflects the potential network demand of each grid area. For example, use a linear combination or a more complex machine learning model to integrate these features. The calculation of the comprehensive feature uses the following formula:

[0173] [Network demand index = α × population density index + β × regional activity intensity]

[0174] Where α and β are coefficients for adjusting the impact of the two features.

[0175] 4. Special event integration.

[0176] Analyze the immediate impact of special events (such as holidays, public events, emergencies, etc.) on network load. By recording the occurrence of these events and analyzing their impact on the load in a specific area, the load forecasting model can be further refined.

[0177] For each event, define an impact radius, which is the geographical range in which it is expected to affect network load. For example, a large music festival may affect network load within a few kilometers.

[0178] Use historical data to analyze the specific impact of events on network load and establish a relationship model between event type, impact level and network load. The model is linear regression, using the following formula:

[0179] [Load i，t =α+β1×Event i +β2×Attendees i +∈ i，t 】

[0180] In the formula, Load i，t is the impact of event i on the network load at time t, Event i is the type of event (such as public events, holidays, etc.), Attendees i is the expected number of participants, α, β1, β2 are model parameters, ∈ i，t is the error term.

[0181] The above four types of feature data (real-time average traffic, traffic differences between business districts on weekdays and weekends, fusion results of demographic data and regional activity patterns, and fusion results of special events) are fused in parallel through the PCA algorithm, and then the principal component data corresponding to the grid is output, which is the second fusion feature mentioned above. This data not only captures the changes in time and space, but also integrates the dynamic information of social and economic activities. Using these features, the model can more accurately predict the changes in base station network load and optimize the configuration and management of network resources.

[0182] On the basis of the above, this application will also extract the terrain characteristics and building characteristics of the candidate grid area to determine the regional site selection suitability (i.e. GIS enhanced grid analysis) as follows:

[0183] In this technical solution, GIS-enhanced raster analysis will be introduced based on the raster features after step fusion. By combining intelligent geographic information system (GIS) with mobile network data (MR), efficient decision support is provided for the optimal positioning of base stations. This method mainly relies on detailed GIS analysis of the geographic area identified by the raster ID, and uses terrain, building configuration, user density and historical network performance data to optimize the layout and performance of base stations. The system collects geographic location information corresponding to each raster ID, including basic geographic attributes such as longitude, latitude, and altitude. Using MR data, the system calculates the network usage of each grid area in real time, including indicators such as the number of users, data transmission volume, and signal strength, such as:

[0184] 1. Terrain enhanced raster analysis.

[0185] The system will calculate the terrain undulation of each grid area in real time and use the digital elevation model (DEM) to estimate the potential impact of terrain on signal propagation. The terrain analysis uses the following formula:

[0186]

[0187] Where, T i represents the average elevation of grid i (i.e., terrain features), h j is the elevation of measurement point j within the grid, and n represents the total number of measurement points.

[0188] 2. Building enhanced raster analysis.

[0189] Considering the impact of buildings on signal blocking, the system calculates signal attenuation in real time by analyzing the layout, height and materials of the building. The real-time calculation formula for signal attenuation by buildings is:

[0190]

[0191] In the formula, A i is the total signal attenuation of grid i (i.e., building characteristics), B k is the attenuation coefficient of the kth building, D k is the distance from the building to the center of the grid.

[0192] It is worth noting that the raster fusion features include not only the first fusion features and the second fusion features, but also the above-mentioned terrain features and building features.

[0193] Accordingly, the regional site selection adaptability of the candidate grid area is calculated by combining the first fusion feature, the second fusion feature, the terrain feature and the building feature, and is used for screening the target grid area. The specific calculation formula can be set by the technicians and will not be described here.

[0194] It can be understood that in this embodiment, by improving the real-time data processing and analysis capabilities, the system can respond to changes in network status more quickly and accurately (such as extracting user behavior and extracting network load characteristics considering factors such as festivals), improving positioning accuracy and real-time network optimization. Secondly, advanced data fusion technology is applied to integrate data sources of more dimensions to achieve more accurate user positioning and network resource allocation. In addition, more advanced machine learning algorithms are introduced, which can handle more complex data relationships and patterns, and provide more accurate load prediction and network configuration optimization.

[0195] In a feasible implementation manner, the step of determining the regional site selection adaptability of the candidate grid area based on the first fusion feature, the second fusion feature, the terrain feature, and the building feature includes steps S261 to S262:

[0196] Step S261, determining the network coverage quality of the candidate grid area through the first fusion feature, the second fusion feature, the terrain feature and the building feature;

[0197] Step S262: performing weighted calculation based on the network coverage quality and the construction cost of building a base station in the candidate grid area to obtain the regional site selection adaptability of the candidate grid area.

[0198] Exemplarily, based on the above-mentioned first fusion feature, the second fusion feature, the terrain feature and the building feature, the network coverage quality of the candidate grid area can be calculated, such as the network coverage quality can be obtained by weighted summation. It is worth noting that the network coverage quality can characterize the quality of network services that can be provided by the base station after the base station is built in the grid in theory. In practical applications, the cost of building a base station should also be considered. Therefore, the regional site selection fitness of the above-mentioned candidate grid area can be obtained by weighted calculation based on the network coverage quality and the construction cost of building a base station in the candidate grid area.

[0199] Correspondingly, considering the cost, the calculation formula of regional location fitness is as follows:

[0200] [F(x)=ω1C(x)+ω2S(x)-ω3O(x)]

[0201] In the formula, F(X) is the fitness function (that is, the regional location fitness mentioned above), C(X) represents the construction cost, S(X) represents the network coverage quality, 0(X) represents other operating costs, ω n is the weight of each factor.

[0202] Correspondingly, in the case of the above calculation formula, the higher the regional site selection fitness, the lower the probability of being selected as the target grid area.

[0203] In addition, when selecting the site for a targeted base station, it is necessary to ensure network coverage and service quality while reducing the construction and maintenance costs of the base station as much as possible. Accordingly, the algorithm formula used is as follows:

[0204] [minimizef(x)=C(x)+λ(1-Q(x))]

[0205] Among them, f(X) is also the fitness function (that is, the fitness of the regional site selection mentioned above, and its output value is made as small as possible). In this formula, C(X) represents the total cost related to base station construction and operation and maintenance, which includes the necessary expenditure costs such as base station building material investment cost, equipment maintenance cost, and human operation and maintenance cost. Q(x) represents the network coverage quality, and λ is the weight coefficient, which is used to adjust the relative importance between cost and coverage quality.

[0206] Correspondingly, in the case of the above calculation formula, the lower the regional site selection fitness, the lower the probability of being selected as the target grid area.

[0207] In a feasible implementation manner, the step of determining the network coverage quality of the candidate grid area by using the first fusion feature, the second fusion feature, the terrain feature, and the building feature includes step S2611:

[0208] Step S2611, respectively normalize the first fusion feature, the second fusion feature, the terrain feature, and the building feature to obtain normalized results, and accumulate the normalized results to obtain the network coverage quality.

[0209] It should be noted that, since the normalization process of each feature is basically the same, in this embodiment, the normalization of each feature can be performed by referring to the following steps.

[0210] An exemplary, normalized calculation formula is as follows:

[0211]

[0212] In the formula, β i is the result after feature i is normalized, that is, the normalized result. i is the weight ratio of feature i, n is the total number of features, x i is the eigenvalue corresponding to feature i.

[0213] Through the above method, each feature can be normalized to obtain each normalized result, and then the normalized results are accumulated to obtain the network coverage quality.

[0214] It is worth noting that in each normalization process of accumulation, normalization can be performed again. For example, the calculation formula of the accumulated network coverage quality is as follows:

[0215]

[0216] Where S(X) is the network coverage quality, β n is the result after feature n is normalized, x n is the eigenvalue corresponding to feature n.

[0217] In a feasible implementation manner, the multidimensional information features include measurement data of a signaling network management system, measurement data of a geographic information system, weather data, and traffic flow data, and the step of fusing the multidimensional information features to obtain a first fused feature of the candidate grid area includes steps S221 to S222:

[0218] Step S221, performing weighted calculation on the measurement data of the signaling network management system, the measurement data of the geographic information system, the weather data and the traffic flow data to obtain the first fusion feature; or

[0219] Step S222: inputting the measurement data of the signaling network management system, the measurement data of the geographic information system, the weather data and the traffic flow data into a preset fusion model to obtain the first fusion feature.

[0220] It should be noted that multi-source data fusion is the process of integrating data from different sources to obtain more comprehensive and accurate analysis results than a single data source. For measurement report (MR) data (such as RSRP and RSRQ) of the signaling network management system, geographic information system (GIS) data (such as terrain, building location, etc.), weather information and traffic flow data, multi-source fusion can greatly enhance the efficiency and effectiveness of network optimization, planning and operation. In this embodiment, two fusion methods are given, namely weighted fusion and model fusion.

[0221] For example, for the weighted fusion method, each data source is assigned different weights according to its reliability and influence. For example, for network optimization, RSRP and RSRQ data may have higher weights than GIS data because they directly reflect the real-time performance of the network. The specific weights can be set by technicians according to actual needs and will not be described here.

[0222] For example, the fusion formula is as follows:

[0223]

[0224] Where V i represents the value of the ith data source (i.e., measurement data from the signaling network management system, measurement data from the geographic information system, weather data, and traffic flow data, etc.), W i Indicates the corresponding weight.

[0225] In another implementation, a fusion model can be used to fuse multi-source data. For example, the features of different data sources, such as measurement data from the signaling network management system (i.e., signal strength of MR data), measurement data from the geographic information system (building height in GIS data), weather data (rainfall, etc.), and traffic flow data are extracted and then fused through a machine learning model neural network for prediction or classification. The specific implementation steps are as follows:

[0226] First, the system extracts key features that are helpful for prediction or classification tasks from various data sources, for example, RSRP and RSRQ from MR data; terrain height and building density from GIS data; temperature and humidity from weather data; and vehicle flow and speed from traffic flow data.

[0227] MR data: Extract key indicators such as RSRP (reference signal received power) and RSRQ (reference signal received quality) from the measurement report of the signaling network management system. These indicators directly reflect the signal quality of the network.

[0228] GIS data: Features such as terrain height and building density are extracted from geographic information systems. This information helps analyze the impact of environmental factors on signal propagation.

[0229] Weather data: Extract meteorological parameters such as temperature and humidity, as these conditions can significantly affect the strength and quality of wireless signals.

[0230] Traffic flow data: including vehicle volume and speed, which is important for analyzing network demand at specific times and locations.

[0231] All extracted features need to be normalized to ensure that the data is analyzed on the same scale and eliminate the bias caused by different measurements:

[0232] Minimum-maximum normalization linearly scales the eigenvalues ​​to the range of 0 to 1, as follows:

[0233]

[0234] In the formula, X man is the largest eigenvalue, X min is the minimum eigenvalue, X is the current eigenvalue, and X` is the result after scaling. After the above scaling, the weight of the data in the model training can be balanced.

[0235] Zscore standardization: transform the data into a distribution with a mean of 0 and a standard deviation of 1. This method is particularly suitable for situations where the data distribution is approximately normal.

[0236] Select a neural network as the training model, configure a multi-layer perceptron (MLP) or a deep neural network, determine the appropriate number of layers and the number of neurons in each layer, and select the activation function ReLU to increase the nonlinear processing capability of the model.

[0237] A known data set can be used for model training, where the known data set is a data set with known fusion results, and the known fusion results can be obtained by manual labeling. In addition, in the model training project, back propagation and gradient descent algorithms can be used to optimize the weights and biases of the model, monitor the loss function (such as mean square error or cross entropy) during the training process, and perform model evaluation and tuning.

[0238] Reference Figure 3 , is a flow chart of a third embodiment of the base station site selection method of the present application, which is proposed based on the first embodiment and the second embodiment of the base station site selection method of the present application. In this embodiment, the same or similar contents as the above embodiments can be referred to the above introduction, and will not be repeated later. The step of determining the location of the target base station according to the target grid area includes steps S31 to S33:

[0239] Step S31, generating each candidate base station location based on the target grid area;

[0240] Step S32, determining the base station site selection fitness of each candidate base station location, and selecting a pending base station location from the candidate base station locations according to the base station site selection fitness;

[0241] Step S33, deriving new candidate base station locations based on the pending base station location, and returning to execute the step of determining the base station site fitness of each candidate base station location based on the new candidate base station locations, until a preset number of iterations is reached, and the final pending base station location is used as the target base station location.

[0242] It should be noted that after determining the target grid area, it is necessary to select a specific base station location from the target grid area. As follows:

[0243] Exemplarily, each candidate base station location is generated based on the target grid area, such as randomly selecting several locations from the target grid area as each candidate base station location. Then determine the base station site fitness of each candidate base station location. It should be noted that the process of determining the base station site fitness of the candidate base station location can refer to the above-mentioned process of generating regional site fitness, only replacing the grid with a specific candidate base station location, so it will not be repeated here. Then, according to the fitness of each base station site selection, the pending base station location is screened out from each candidate base station location, such as the candidate base station location with the highest base station site fitness can be used as the pending base station location. Then derive new candidate base station locations from the pending base station location, wherein the derivation process may include fitting to obtain the direction of increasing base station site fitness, extending the increasing direction on the basis of the pending base station location, and advancing one step to obtain a new candidate base station location. In actual applications, different increasing directions may be fitted, and correspondingly, different new candidate base station locations may be obtained. After deriving the new candidate base station locations, return to execute the step of determining the base station site selection fitness of each candidate base station location based on the new candidate base station locations, until the preset number of iterations is reached, and then use the final pending base station location as the above-mentioned target base station location. At this point, the site selection of the base station is completed.

[0244] Reference Figure 4 , is a flow chart of a fourth embodiment of the base station site selection method of the present application, which is proposed based on the first embodiment, the second embodiment, and the third embodiment of the base station site selection method of the present application. After the step of determining the location of the target base station according to the target grid area, the method further includes steps S01 to S02:

[0245] Step S01, determining the signal quality information of the base station after the base station is built at the target base station location;

[0246] Step S02: Optimize the operating parameters of the base station built at the target base station location based on the quality information.

[0247] It should be noted that after completing the site selection of the base station, in this embodiment, the base station will be further estimated and optimized to optimize the resource configuration of the base station.

[0248] Exemplarily, the signal quality information of the base station after the base station is built at the target base station location is determined. That is, adaptive signal strength evaluation, by analyzing the signal strength information in the signaling network management system measurement report (MR), realizes the dynamic adjustment of the signal strength model, thereby optimizing network performance and user experience.

[0249] like:

[0250] MR data is collected from the signaling network management platform, which includes the signal strength information of each signaling network management system in each grid ID, such as reference signal received power (RSRP) and reference signal received quality (RSRQ).

[0251] A signal strength prediction model is established using statistical analysis and machine learning methods. The model analyzes the average signal strength within the grid based on historical data and predicts possible signal attenuation or enhancement.

[0252] The modeling formula is as follows:

[0253] [Predicted i =α+β1×Obstacle i +β2×Distance i +∈ i ]

[0254] Among them, Predicted i is the predicted signal receiving power (i.e., the signal quality information of the above base station), i Indicates the impact of obstacles, Distance i is the distance to the nearest base station, α, β1, β2 are model parameters, ∈ t is the error term.

[0255] Based on the output of the signal strength model, the base station's transmit power and antenna direction are dynamically adjusted. The adjustment algorithm takes into account the user's geographical distribution and real-time traffic data to optimize signal coverage and reduce interference.

[0256] The formula used in the adjustment process is as follows:

[0257] [Adjusted i =Base+Y×(Target-Predicted i )]

[0258] Among them, Adjusted i is the adjusted transmit power (i.e., the base station operating parameter), Predicted i is the predicted signal receiving power, Base is the basic transmission power of the base station, Target is the target signal receiving power, and λ is the adjustment coefficient.

[0259] After implementing signal strength adjustments, continuously monitor network performance indicators such as user satisfaction, call drop rate, and data transmission speed to evaluate the effectiveness of the adjustments. Based on performance feedback data, further optimize the signal strength model and adjustment strategy to ensure continuous improvement in network service quality.

[0260] In addition, for completed base stations, a dynamic resource allocation decision algorithm can also be used to dynamically adjust resource allocation, such as spectrum resources and power, based on network load prediction and user distribution models to optimize network performance.

[0261] The time series model is built using historical MR data to predict the network load of each grid. The machine learning method Gradient Boosting Tree (GBT) is used to achieve accurate prediction.

[0262] The forecasting model takes into account factors such as daily cycles, weekly cycles, and seasonal changes. Especially during large events or holidays, the model needs to be able to accurately predict the surge in network load. The load forecasting formula is as follows:

[0263] [Load t+1 =f(Load t , Time, Event, User)]

[0264] Where f represents the prediction function, Load t It is the network load in the current period, Time includes time-related factors, Event is event information, and User is the user behavior pattern.

[0265] According to the load forecast results, the spectrum allocation and base station transmission power of each grid area are dynamically adjusted. This process involves a multi-objective optimization problem aimed at maximizing network coverage and throughput while minimizing interference and cost.

[0266] The spectrum and power allocation strategy uses optimization algorithms, such as linear programming or genetic algorithms, to ensure the reasonable allocation of resources among the grids. The formula is as follows:

[0267]

[0268] In the formula, C i (X i ) indicates that the i-th grid is in the distribution of X i resources, and X is the total available resources.

[0269] The real-time adjustment algorithm quickly adjusts spectrum allocation based on real-time traffic data and user feedback to respond to emergencies or unexpected load increases.

[0270] In addition, it is worth mentioning that in the process of selecting a base station site, the site selection process can be visualized. Specifically:

[0271] The base station site selection visualization technology based on grid ID intelligent MR data fusion and targeted base station optimization positioning algorithm combines advanced geographic information system (GIS) technology and real-time data processing to provide a dynamic visualization tool to optimize base station site selection and network performance management. This technology is achieved through the following key steps:

[0272] The collected MR data are encoded according to geographic location using GIS technology, and each grid ID represents a specific area. These data include but are not limited to signal strength, number of users, data traffic, etc.

[0273] On the visual map, different network statuses are represented by different colors or icon sizes, allowing network operators to intuitively identify the quality of network coverage and any possible service hotspots or shortcomings.

[0274] Accordingly, you can set up a variety of chart types (such as line charts, bar charts, and heat maps) to display data about time series analysis or performance comparisons between regions. Heat maps can be used to represent traffic density or signal quality in different areas, helping to analyze which areas may need additional network support or base station optimization during a specific time period.

[0275] The visualization interface supports real-time data updates to ensure that the displayed information can reflect the real-time status of the network. This requires an efficient and reliable data synchronization mechanism between the front-end visualization components and the back-end data processing system.

[0276] Ensure visualizations provide clear and accurate information at different zoom levels, using vector graphics and adaptive layout designs to keep data clear and readable at a variety of display sizes.

[0277] Provides an intuitive user interface that allows operators to zoom in on specific areas and filter specific types of data by simply clicking and dragging, enhancing user experience and improving operational efficiency.

[0278] The present application provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the base station site selection method in the above-mentioned embodiment one.

[0279] Reference below Figure 5, which shows a schematic diagram of the structure of an electronic device suitable for implementing the embodiment of the present application. The electronic device in the embodiment of the present application may include but is not limited to mobile terminals such as computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), and fixed terminals such as computers. Figure 5 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0280] like Figure 5 As shown, the electronic device may include a processing device 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the electronic device are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows an electronic device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0281] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0282] The electronic device provided by the present application adopts the base station site selection method in the above embodiment, which can solve the technical problem that the accuracy of base station site selection in the related technology is not high and it is difficult to ensure that the optimal base station location can be determined. Compared with the prior art, the beneficial effects of the electronic device provided by the present application are the same as the beneficial effects of the base station site selection method provided by the above embodiment, and other technical features in the electronic device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0283] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0284] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0285] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the base station site selection method in the above-mentioned embodiment.

[0286] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0287] The computer-readable storage medium may be included in the electronic device, or may exist independently without being installed in the electronic device.

[0288] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by an electronic device, the electronic device:

[0289] Divide the target area into candidate grid areas;

[0290] For any one of the candidate grid regions, determining the regional site selection adaptability of the candidate grid region based on the grid fusion features of the candidate grid region;

[0291] After traversing the candidate grid areas, based on the regional site selection adaptability of the candidate grid areas, a target grid area is screened out from the candidate grid areas, and the target base station position is determined according to the target grid area.

[0292] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0293] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0294] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0295] The readable storage medium provided by the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned base station site selection method, and can solve the technical problem that the accuracy of base station site selection in the related art is not high and it is difficult to ensure that the optimal base station location can be determined. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as the beneficial effects of the base station site selection method provided by the above-mentioned embodiment, and will not be repeated here.

[0296] The present application also provides a computer program product, including a computer program, which implements the steps of the base station site selection method as described above when executed by a processor.

[0297] The computer program product provided in this application can solve the technical problem of base station site selection. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the base station site selection method provided in the above embodiment, which will not be repeated here.

[0298] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A base station site selection method, characterized in that: The base station site selection method comprises the following steps: Divide the target area into candidate grid areas; For any one of the candidate grid regions, determining the regional site selection adaptability of the candidate grid region based on the grid fusion features of the candidate grid region; After traversing the candidate grid areas, based on the regional site selection adaptability of the candidate grid areas, a target grid area is screened out from the candidate grid areas, and the target base station position is determined according to the target grid area.

2. The base station site selection method according to claim 1, characterized in that: The grid fusion feature includes a first fusion feature and a second fusion feature. The step of determining the regional site selection adaptability of the candidate grid area based on the grid fusion feature of the candidate grid area includes: Extracting multi-dimensional information features of the candidate grid area; Fusing the multi-dimensional information features to obtain a first fused feature of the candidate grid region; Extracting user behavior information and network load characteristics corresponding to the candidate grid area; Performing feature fusion based on the user behavior information and the network load feature to obtain a second fusion feature; Extracting terrain features and building features of the candidate grid area; The regional site selection adaptability of the candidate grid area is determined based on the first fusion feature, the second fusion feature, the terrain feature, and the building feature.

3. The base station site selection method according to claim 2, characterized in that: The step of determining the regional site selection adaptability of the candidate grid area based on the first fusion feature, the second fusion feature, the terrain feature, and the building feature comprises: Determining the network coverage quality of the candidate grid area by using the first fusion feature, the second fusion feature, the terrain feature, and the building feature; The regional site selection adaptability of the candidate grid area is obtained by performing a weighted calculation based on the network coverage quality and the construction cost of building a base station in the candidate grid area.

4. The base station site selection method according to claim 3, characterized in that: The step of determining the network coverage quality of the candidate grid area by using the first fusion feature, the second fusion feature, the terrain feature, and the building feature comprises: Normalization processing is performed on the first fusion feature, the second fusion feature, the terrain feature, and the building feature respectively to obtain normalized results, and the normalized results are accumulated to obtain the network coverage quality.

5. The base station site selection method according to claim 2, characterized in that: The multi-dimensional information features include measurement data of a signaling network management system, measurement data of a geographic information system, weather data, and traffic flow data. The step of fusing the multi-dimensional information features to obtain a first fused feature of the candidate grid area includes: Performing weighted calculation on the measurement data of the signaling network management system, the measurement data of the geographic information system, the weather data, and the traffic flow data to obtain the first fusion feature; or The measurement data of the signaling network management system, the measurement data of the geographic information system, the weather data and the traffic flow data are input into a preset fusion model to obtain the first fusion feature.

6. The base station site selection method according to claim 1, characterized in that: The step of determining the location of the target base station according to the target grid area comprises: generating each candidate base station location based on the target grid area; Determining the base station site selection fitness of each candidate base station location, and selecting a pending base station location from the candidate base station locations according to the base station site selection fitness; Based on the pending base station position, new candidate base station positions are derived, and based on the new candidate base station positions, the step of determining the base station site fitness of each candidate base station position is returned to execute until a preset number of iterations is reached, and the final pending base station position is used as the target base station position.

7. The base station site selection method according to claim 1, characterized in that: After the step of determining the location of the target base station according to the target grid area, the method further includes: Determining signal quality information of the base station after the base station is built at the target base station location; The operating parameters of the base station built at the target base station location are optimized based on the quality information.

8. An electronic device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of base station site selection according to any one of claims 1 to 7.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of base station site selection according to any one of claims 1 to 7 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of selecting a base station site according to any one of claims 1 to 7 are implemented.

Images