A method for analyzing new site business planning based on a live network topology

Abstract

The application discloses a kind of based on present network topology's new site service planning analysis method, comprising: present network is carried out topological data collection, obtains present network topological data and service data;Utilize planning analysis model to process the data, build present network topological relationship model and identify service planning demand area;According to the demand area and topological relationship model, new site site selection planning and service configuration prediction are carried out;According to planning result, generate optimization instruction, control network management system to execute new site deployment and service adjustment action.The present application realizes the accurate identification of covering weak point and high load area through present network topological relationship model, combined with optimization algorithm, completes the automatic planning of new site position and service configuration, and introduces the real-time feedback mechanism after execution Dynamic correction model parameters, effectively improve the accuracy, efficiency and service shunting effect of network planning.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and more specifically to a method for planning and analyzing new site services based on existing network topology. Background Technology

[0002] With the rapid development of mobile communication technology and the continuous growth of user service demands, the topology of existing base stations is becoming increasingly complex, and the distribution of service traffic exhibits significant regional differences. Operators need to continuously add sites to improve network coverage quality and service carrying capacity. Traditional site addition processes often rely on engineers' personal experience or rough judgments of local coverage gaps. While this approach can solve some urgent problems in the short term, it lacks a systematic, forward-looking, and holistic perspective in the long run. With the popularization of 5G, IoT, and edge computing, service types have expanded from simple person-to-person communication to the Internet of Everything, fundamentally changing the network traffic model. If adding sites only considers coverage gap filling, it is highly likely to lead to resource waste, increased interference, or service bottlenecks.

[0003] Current site planning methods are typically based on historical call statistics or simple propagation model predictions, failing to fully utilize the rich information implicit in the existing network topology, failing to effectively integrate existing network topology data with real-time service data, and failing to accurately identify areas with weak coverage and high service loads. At the same time, they lack dynamic feedback and optimization mechanisms for planning results, leading to problems such as low efficiency in deploying new sites, poor service offloading effects, and waste of network resources. In addition, in traditional planning processes, service planning and analysis stages usually lag behind site selection design, lacking quantitative evaluation of the synergistic effects between new sites and existing network nodes, making it difficult to achieve the optimal balance between return on investment and user experience.

[0004] Therefore, the industry urgently needs a new site service planning and analysis method that can integrate the existing network topology, historical service data, and geographic information. A search revealed a Chinese invention patent with application number 202311509750.7, which discloses a communication site planning method based on energy big data. The method includes the following steps: acquiring resource allocation data for substations of different voltage levels in different regions and analyzing key factors affecting resource allocation to obtain a substation data sample set; using a Gaussian mixture model (GMM) to determine the probability value of each substation data sample belonging to each cluster and clustering them into the cluster with the highest probability value, determining the site resource critical threshold, and obtaining a standard for classifying and grading different communication substation sites; when classifying planned substation sites, performing LightGBM classification on the substation site data samples to be classified and the clustering indicators to obtain the classification level of the planned substation sites; and using a data-driven method to achieve medium- and long-term high-reliability communication network perception and prediction under conditions lacking power grid planning topology. The patented technical solution combines two machine learning models, GMM and LightGBM. The decision-making process of the LightGBM model is relatively complex, making it difficult to intuitively explain to planners "why a certain site is classified into a specific level". The application of the model is also not flexible enough. Summary of the Invention

[0005] In response to the problems in power communication site planning mentioned in the background technology, this invention proposes a new site service planning and analysis method based on the existing network topology. Through deep integration and intelligent analysis of existing network topology data and service data, an accurate existing network topology relationship model is constructed, which can effectively identify areas with weak network coverage and areas with high service load, and realize intelligent and automated prediction of new site site selection planning and service configuration.

[0006] A method for planning and analyzing new site services based on existing network topology includes the following steps: S1: Collect topology data from the existing network to obtain network topology data and service data; S2: Use the planning analysis model to process the existing network topology data and business data, construct the existing network topology relationship model and identify the business planning requirement areas; S3: Based on the business planning requirements area and the existing network topology model, perform new site selection planning and business configuration prediction; S4: Generate optimization instructions based on the planning results, and control the network management system to execute new site deployments and service adjustments.

[0007] The specific steps of S1 are as follows: S101: Obtain the geographical location information, cell parameters, and neighbor cell relationships of existing base stations from the network management system; S102: Collect live network service data, including traffic statistics, user distribution, and performance metrics. S103: Clean, correlate, and standardize the acquired data to obtain standardized live network topology data and business datasets.

[0008] The specific steps of S2 are as follows: S201: Extract features from the standardized network topology data and service data to obtain topology features and service features; S202: Input the aforementioned features into the planning analysis model to construct a weighted topology graph model; S203: Identify areas with weak coverage and areas with high service load based on the topology model, and use these as areas for service planning requirements.

[0009] Specifically, before executing step S2, it is necessary to train the parameters of the planning and analysis model based on historical business data and current network topology data; the planning and analysis model is used to comprehensively consider topology connectivity, interference relationships and business growth trends.

[0010] The specific steps of S3 are as follows: S301: Determine a set of candidate site locations within the business planning requirement area; S302: Utilize optimization algorithms combined with the existing network topology model to select the optimal location and service configuration parameters for new sites; S303: Predicts the service offloading effect and network performance improvement indicators after the deployment of new sites.

[0011] The specific steps of S4 are as follows: S401: Convert the planning results into new site deployment instructions, business parameter configuration instructions, and resource adjustment instructions; S402: Use the optimization instructions to drive the network management system to perform real-time deployment of new sites and service adjustments; S403: During execution, the existing network topology data is continuously fed back, the input of the planning analysis model is dynamically updated, and the optimization scheme is regenerated.

[0012] Specifically, the process of cleaning, associating, and standardizing the acquired data described in S103 involves spatially associating base station location information with geographic information system data, constructing an adjacency matrix based on neighboring cell relationships, and performing time and spatial aggregation processing on the business data to generate gridded business distribution data.

[0013] Specifically, in S203, the identification of weak coverage areas and high service load areas based on the topology graph model involves calculating the topology centrality and service load index of each node, comparing the calculation results with preset coverage thresholds and capacity thresholds, and identifying areas that do not meet the thresholds as service planning requirement areas.

[0014] Specifically, the specific operation of S302 is as follows: construct a planning cost map that includes the current network topology relationship model and the service planning requirement area; use heuristic search or linear programming methods to search for the optimal site deployment scheme on the planning cost map; perform service simulation verification on the selected deployment scheme to obtain the optimized service configuration parameters.

[0015] Specifically, the specific operation of S403 is as follows: real-time monitoring of key performance indicators after the deployment of new sites, and comparison with the predicted network performance improvement indicators; calculation of the deviation between the actual service diversion effect and the predicted effect; when the deviation exceeds a preset threshold, automatic triggering of the planning analysis model to correct parameters, and re-execution of new site location planning or service configuration adjustment.

[0016] The advantages and positive effects of this invention are: 1. The new site service planning and analysis method disclosed in this invention constructs a high-fidelity network topology model by cleaning, associating and standardizing heterogeneous data such as base station geographical location, neighbor cell relationship and service traffic, which effectively avoids resource waste and planning errors.

[0017] 2. The new site business planning analysis method disclosed in this invention automatically identifies business planning demand areas through a planning analysis model. The model comprehensively considers topological connectivity, interference relationships, and business growth trends, and can accurately calculate areas with weak coverage and high business load. This mechanism of automatically extracting planning targets from massive amounts of data not only significantly reduces the workload of manual surveying and analysis, but also ensures the objectivity and comprehensiveness of planning demand identification, making it particularly suitable for complex and ever-changing environments.

[0018] 3. The new site business planning and analysis method disclosed in this invention introduces a site selection planning process that combines optimization algorithms with simulation verification. By performing heuristic search or linear programming on a planning cost map, and supplemented by business simulation verification, the optimal location of the new site and its business configuration parameters can be quickly selected. Compared with traditional manual site selection, this automated decision-making method significantly improves planning efficiency and can effectively predict the business diversion effect and performance improvement indicators after the new site is deployed, providing reliable data support for investment decisions.

[0019] 4. The new site service planning and analysis method disclosed in this invention constructs a closed-loop automated execution and feedback optimization system. By converting the planning results into executable instructions for the network management system, it achieves automated connection from planning to deployment, enabling real-time monitoring of the actual performance of new sites after deployment and comparison with predicted values. When a large deviation occurs, it automatically triggers model parameter correction and regenerates the optimization scheme, thereby ensuring that network planning can dynamically adapt to business changes and environmental impacts, achieving continuous self-optimization. Attached Figure Description

[0020] Figure 1 A flowchart illustrating the new site service planning and analysis method based on the existing network topology provided by this invention; Figure 2 The flowchart of the method for processing existing network data using a planning analysis model provided by the present invention is shown below; Figure 3 The flowchart illustrates the new site selection planning and service configuration prediction method provided by this invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be described in more detail below with reference to the accompanying drawings. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are some, but not all, embodiments of this invention. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this invention, and should not be construed as limiting the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0022] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0023] To facilitate understanding of the embodiments of the present invention, the following will provide further explanation and description with reference to the accompanying drawings and several specific embodiments. These embodiments do not constitute a limitation on the embodiments of the present invention.

[0024] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0025] like Figure 1As shown, the method for planning and analyzing new site services based on the existing network topology includes the following steps: S1: Collect topology data from the existing network to obtain network topology data and service data; In this embodiment, this step establishes a standardized data communication interface with the network management system to automate the collection of existing network topology and service data. The collection process can employ scheduled tasks or event-triggered mechanisms to obtain topology-related information from the network management system, such as the geographical location information of existing base stations, cell parameter configurations, and neighbor cell relationships. Simultaneously, it acquires service data such as traffic statistics, user distribution, and performance indicators for each cell at different times. The collected raw data then undergoes preprocessing operations such as integrity verification, missing value completion, anomaly filtering, and format conversion, ultimately forming structured and standardized existing network topology and service data, providing accurate and reliable basic input for subsequent model processing.

[0026] S2: Use the planning analysis model to process the existing network topology data and business data, construct the existing network topology relationship model and identify the business planning requirement areas; After receiving the aforementioned existing network topology and service data, the planning and analysis model first performs multi-dimensional feature extraction and correlation analysis to obtain key information reflecting the connectivity relationships between base stations and the distribution of services. Subsequently, the model constructs a network topology model containing weighted information, using base stations as nodes and neighboring cell relationships as edges. Based on this, the model automatically identifies service planning demand areas requiring new site planning by comprehensively calculating the topological connectivity and service load levels of all regions in the network, ensuring that subsequent planning work is clearly targeted. Understandably, this process fully utilizes the inherent correlation of real-time network data, making the topology model output by the model highly consistent with the actual network conditions.

[0027] S3: Based on the business planning requirements area and the existing network topology model, perform new site selection planning and business configuration prediction; In this embodiment, this step focuses on the identified service planning demand area and combines it with the existing network topology model to plan the location of new sites. First, multiple candidate site location sets are generated within the demand area. Then, an optimization algorithm is used to comprehensively consider factors such as topological constraints, coverage overlap, and service offloading potential to search and determine the optimal location for the new site. Simultaneously, the model also predicts the service configuration of the new site at the selected location, including quantitative results such as the predicted service carrying capacity and network performance improvement indicators after deployment, providing scientific and actionable decision support for actual deployment.

[0028] S4: Generate optimization instructions based on the planning results, and control the network management system to execute new site deployments and service adjustments.

[0029] In this embodiment, the system converts the optimal planning result into optimization instructions that the network management system can directly recognize and execute. These optimization instructions include detailed information such as deployment parameters for new sites, service configuration schemes, and resource adjustment strategies. After the optimization instructions are sent to the network management system through a standardized northbound interface, the network management system drives the corresponding network elements to complete the activation of new sites, parameter configuration, and dynamic adjustment of network services. This ensures that the planning scheme can be quickly and accurately implemented in the existing network environment and achieve its optimization effect.

[0030] The specific steps of S1 are as follows: S101: Obtain the geographical location information, cell parameters, and neighbor cell relationships of existing base stations from the network management system; In this embodiment, this step establishes a secure, standardized northbound interface with the network management system to automate the collection of existing base station topology information. The system can extract geographical location information such as latitude and longitude coordinates, station height, and antenna azimuth angle for each base station, as well as cell parameters and neighbor cell relationship data, such as cell ID, frequency configuration, transmit power, and neighbor cell list, from the network management system using either periodic polling or event triggering. This data directly constitutes the core nodes and connections of the existing network topology, providing complete foundational information for the subsequent construction of the topology relationship model.

[0031] S102: Collect live network service data, including traffic statistics, user distribution, and performance metrics. Furthermore, this step simultaneously involves collecting live network service data. The system obtains traffic statistics, user distribution density information, and key performance indicators (such as uplink and downlink throughput, call drop rate, handover success rate, and interference level) for each cell within a specified time granularity from the network management system or associated performance management system. This service data, together with the topology data, reflects the actual operating status and service load distribution of the live network, providing necessary quantitative support for subsequently identifying areas with service planning needs.

[0032] S103: Clean, correlate, and standardize the acquired data to obtain standardized live network topology data and business datasets.

[0033] Specifically, the system performs unified cleaning, association, and standardization processing on the collected topology and business data. This includes cleaning operations such as removing duplicate records, filling in missing values, and filtering outlier data; association operations such as spatially associating base station geographic location information with geographic information system data and automatically constructing adjacency matrices based on neighboring cell relationships; and standardization operations such as unifying data formats, normalizing numerical ranges, and generating gridded business distribution data. After processing, structured and standardized live network topology and business datasets are obtained, which can be directly used as input to the planning and analysis model, ensuring efficient and accurate subsequent processing.

[0034] In this step, the steps of cleaning, correlating, and standardizing the acquired data include: Spatial association between base station location information and geographic information system data; Construct an adjacency matrix based on neighbor relationships; Perform time and spatial aggregation processing on business data to generate gridded business distribution data.

[0035] The specific steps of S2 are as follows: S201: Extract features from the standardized network topology data and service data to obtain topology features and service features; In this embodiment, this step first receives standardized live network topology data and service datasets obtained from previous processing. The system extracts topological features from the topology data, such as node degree, betweenness centrality, clustering coefficient, and connection strength in the adjacency matrix. Simultaneously, it extracts service features from the service data, such as traffic density, user activity, peak throughput, and performance trend for each grid or base station. The feature extraction process employs multi-dimensional statistical and vector representation methods to transform the raw structured data into quantifiable high-dimensional feature vectors, ensuring that subsequent models can comprehensively capture the topological characteristics and service distribution patterns of the live network.

[0036] S202: Input the aforementioned features into the planning analysis model to construct a weighted topology graph model; The system takes the extracted topological features and service feature vectors as input and feeds them into the planning and analysis model for fusion calculation. The planning and analysis model automatically constructs a weighted topology graph model with multiple weighted attributes, using base stations or grids as nodes and extracted neighbor cell relationships and weights as edges. The weights comprehensively consider dimensions such as topological connectivity, geographical distance, service load, and potential interference factors. The features are iteratively updated using graph neural networks or graph embedding algorithms, ultimately generating a weighted topology graph model that can fully represent the overall topological relationships and service distribution of the current network.

[0037] S203: Identify areas with weak coverage and areas with high service load based on the topology model, and use these as areas for service planning requirements.

[0038] The system performs region identification based on the constructed weighted topology graph model. The model traverses each node and subgraph in the graph, calculating the topological centrality index, coverage integrity score, and comprehensive service load index for each region, and comparing these scores with preset coverage and capacity thresholds. Regions with insufficient topological connectivity, coverage scores below the threshold, or service loads exceeding the capacity threshold are automatically marked as weak coverage regions and high service load regions, and these regions are uniformly identified as service planning requirement regions. The identification results are output in the form of a geographic grid or coordinate list, providing a clear target scope for subsequent planning steps.

[0039] Before executing step S2, it is necessary to train the parameters of the planning and analysis model based on historical business data and current network topology data; the planning and analysis model is used to comprehensively consider topology connectivity, interference relationships and business growth trends.

[0040] In this embodiment, the first step involves collecting historical service data and historical network topology data from the network management system and historical performance database for multiple consecutive periods (typically the past 6-24 months). This includes base station topology structures, neighbor cell relationship change records, traffic distribution data, user distribution, performance index data, and known records of new site deployment effects for different time periods. This historical data is divided into training, validation, and test sets, and undergoes feature extraction processing consistent with real-time data to obtain topological feature vectors (node ​​degree, betweenness centrality, clustering coefficient, etc.) and service feature vectors (traffic density, user activity, performance index trends, etc.). The planning and analysis model adopts a graph neural network (GNN) based architecture, specifically composed of multiple layers of graph convolutional network (GCN) layers, used to process the topology graph structure data. It also integrates fully connected layers and a temporal attention mechanism to capture service growth trends. During training, historical topology features and service features are used as input, while historically known service planning demand area labels or actual deployed service diversion effect indicators are used as supervised output targets. A cross-entropy loss function combined with mean squared error loss is used for joint optimization. The Adam optimizer is used to iteratively update parameters with a batch size of 64, a learning rate of 0.001, and a training duration of 200-500 epochs, until the region recognition accuracy and prediction error on the validation set reach the preset convergence criteria. After training, all model parameters (including the weight matrix of the graph convolutional layer, the parameters of the attention mechanism, and the bias terms of the fully connected layer) are permanently saved.

[0041] The specific steps of S3 are as follows: S301: Determine a set of candidate site locations within the business planning requirement area; Specifically, based on the geographical scope and gridded business distribution data of the identified business planning requirement area, the system first performs uniform sampling within the area at a preset resolution (e.g., a 50m x 50m grid) to generate a large number of potential candidate locations for new sites. Subsequently, the system combines geographic information system data and the existing network topology model to automatically filter out unsuitable candidate points, such as those with excessive overlap with existing base station coverage, topological connectivity conflicts, geographically limited locations, or those with severe potential interference. This results in a set of candidate site locations containing dozens to hundreds of feasible candidate sites. This set directly serves as the search space for subsequent optimization algorithms, ensuring that all candidate locations are within the business planning requirement area and possess practical deployment feasibility.

[0042] S302: Utilize optimization algorithms combined with the existing network topology model to select the optimal location and service configuration parameters for new sites; In this embodiment, a genetic algorithm is used as the optimization algorithm, closely integrated with the existing network topology model to achieve global optimality. The specific process is as follows: First, each candidate site location and its corresponding service configuration parameters (including operating frequency, transmit power, antenna tilt angle, azimuth angle, etc.) are encoded into a chromosome, forming an initial population. Then, a fitness function is defined, which, based on the existing network topology model, comprehensively calculates multiple objectives such as coverage integrity score, service offloading capacity score, topology connectivity score, and interference penalty. The coverage score is calculated based on the node coverage relationship and service load distribution in the model, the service offloading score is based on the traffic migration volume predicted by the model, and the interference penalty is quantified based on the interference relationship matrix in the model. Next, the population is iteratively evolved through genetic operations such as roulette wheel selection, single-point crossover, and uniform mutation. After a preset number of iterations (typically 200 to 500) or after fitness convergence, the individual with the highest fitness is selected as the optimal solution, thereby determining the final optimal new site location and the corresponding optimal service configuration parameters. This process ensures that the optimization result highly matches the existing network topology model and is globally optimal.

[0043] S303: Predicts the service offloading effect and network performance improvement indicators after the deployment of new sites.

[0044] The system performs simulation predictions based on the selected optimal location of new sites and service configuration parameters, combined with the existing network topology model. By establishing a traffic diversion simulation model, it calculates the service diversion ratio, user migration number, and overall network capacity improvement after the deployment of new sites in the target area and surrounding cells. Simultaneously, it quantitatively evaluates key network performance improvement indicators such as coverage improvement percentage, average user throughput increase, call drop rate, and handover success rate optimization. These prediction results are presented in a quantitative report, providing network administrators with an assessment of expected effects before deployment, ensuring that the planning scheme's effectiveness and rationality can be verified before actual implementation.

[0045] The specific steps of S4 are as follows: S401: Convert the planning results into new site deployment instructions, business parameter configuration instructions, and resource adjustment instructions; In this embodiment, this step receives planning results from the preceding planning module, including the optimal location of new sites, service configuration parameters, and resource adjustment suggestions. The system first standardizes and maps these results, converting the abstract planning data into structured instructions that the network management system can directly recognize. Specifically, this includes generating new site deployment instructions (containing parameters such as site ID, latitude and longitude, equipment model, and activation time), service parameter configuration instructions (containing configuration items such as cell frequency, transmit power, antenna tilt angle, azimuth angle, and neighbor cell relationship update list), and resource adjustment instructions (containing carrier resource allocation, transmission bandwidth adjustment, and interference coordination strategies). The conversion process uses a predefined instruction template library and parameter mapping rules to ensure that each instruction format conforms to the network management system's northbound interface specification, while also including execution priority and rollback strategy information, ultimately forming a complete and directly deployable set of optimized instructions.

[0046] S402: Use the optimization instructions to drive the network management system to perform real-time deployment of new sites and service adjustments; Furthermore, the system distributes the aforementioned optimized instruction set to the network management system via a standardized northbound interface. Upon receiving the instructions, the network management system automatically parses and executes them sequentially: first, it drives the hardware activation process for new sites, including equipment power-on, base station registration, and transmission link establishment; then, it batch-distributes service parameter configuration instructions to complete operations such as cell parameter activation, neighbor cell relationship synchronization updates, and resource allocation adjustments; simultaneously, it executes resource adjustment instructions to achieve dynamic load balancing across the entire network. The entire execution process supports real-time monitoring, allowing instructions to be distributed in batches or in parallel based on priority, ensuring that new sites complete the transition from planning to live network operation in the shortest possible time, and providing real-time feedback on the execution status to the planning system.

[0047] S403: During execution, the existing network topology data is continuously fed back, the input of the planning analysis model is dynamically updated, and the optimization scheme is regenerated.

[0048] During the deployment of new sites and the execution of business adjustments, the system continuously collects and provides feedback on existing network topology and business data. Specifically, it automatically retrieves the latest topology, traffic, and performance data at fixed time intervals (typically 5-15 minutes) or when critical events occur (such as site status changes or performance anomalies) through the network management system interface. This real-time feedback data is directly injected into the planning and analysis model, serving as new input for dynamic updates. Based on the feedback results, the model recalculates the topology relationship model, assesses changes in the current business planning requirements area, and triggers optimization algorithms to regenerate optimization solutions. If a significant deviation is detected between the actual execution effect and the planning prediction, the system automatically generates corrected optimization instructions, forming a closed-loop iterative optimization to ensure that the entire new site business planning process remains in an optimal state.

[0049] This embodiment constructs a complete closed-loop iterative optimization mechanism through three core steps: real-time monitoring of key performance indicators after the deployment of new sites and comparison with predicted values; accurate calculation of the deviation between the actual business diversion effect and the predicted effect; and automatic triggering of model parameter correction and replanning when the deviation exceeds a preset threshold. It can promptly detect the deviation between planning and actual execution and make adaptive corrections, ensuring that the deployment effect of new sites continues to approach the optimal state, and significantly improving the long-term accuracy and adaptability of network planning.

[0050] Specifically, key performance indicators after the deployment of new sites are monitored in real time and compared with the predicted network performance improvement indicators.

[0051] In this embodiment, the real-time monitoring process is initiated immediately after the new site deployment command is issued. The system automatically collects key performance indicators (KPIs) of the new site and its surrounding area through the network management system's northbound interface at fixed time intervals (e.g., every 5 minutes) or through event triggering. These KPIs include average uplink and downlink throughput, user access success rate, call drop rate, handover success rate, coverage metrics, and traffic offloading ratio. Simultaneously, the system retrieves network performance improvement indicators (such as coverage improvement percentage and expected throughput growth) predicted before this deployment from the planning results database. The monitored actual and predicted KPIs are mapped to the same quantitative dimension and compared in real time, forming a comparison record table containing actual values, predicted values, and differences. This ensures that structured comparison data is generated for each monitoring cycle, providing direct input for subsequent deviation calculations.

[0052] Furthermore, the deviation between the actual business diversion effect and the predicted effect is calculated.

[0053] The system quantifies the actual traffic diversion effect data (such as the actual traffic diversion percentage of newly added sites, the number of users migrated, and the load reduction percentage of surrounding communities) in the comparison record table with the traffic diversion effect data predicted during the planning phase. Specifically, a relative deviation calculation method is used: the deviation percentage is calculated for each indicator, i.e., deviation value = (actual value - predicted value) / predicted value × 100%, and a weighted average of multiple indicators is obtained to obtain an overall deviation value. Simultaneously, the system also statistically analyzes the cumulative deviation trend over multiple consecutive monitoring periods, forming a time-series deviation curve. This entire calculation process is automatically completed by the deviation calculation module built into the planning system, and the results are written to the feedback database in real time, ensuring that the deviation value calculation is objective, accurate, and traceable.

[0054] When the deviation value exceeds the preset threshold, the planning analysis model is automatically triggered to correct the parameters and re-execute the new site selection planning or business configuration adjustment.

[0055] Understandably, when the calculated overall deviation value or the deviation value of any key indicator exceeds a preset threshold (e.g., overall deviation exceeds 15% or single performance indicator deviation exceeds 10%), the system immediately and automatically triggers an alarm mechanism. The planning and analysis model receives the latest collected network topology data, business data, and deviation feedback records as supplementary inputs to perform supervised parameter correction on the model's internal parameters. For example, it fine-tunes the topology feature weights, business growth trend coefficients, or weight coefficients in the interference relationship matrix through gradient descent. After correction, the model automatically re-executes the new site selection planning or business configuration adjustment process, generating a corrected optimization scheme and a new set of optimization instructions. This triggering and correction process is fully automated, requiring no manual intervention, and supports multiple iterations until the deviation value falls back to the threshold range or reaches the maximum number of iterations, thereby achieving dynamic closed-loop optimization of planning execution.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for planning and analyzing new site services based on existing network topology, characterized in that, Includes the following steps: S1: Collect topology data from the existing network to obtain network topology data and service data; S2: Use the planning analysis model to process the existing network topology data and business data, construct the existing network topology relationship model and identify the business planning requirement areas; S3: Based on the business planning requirements area and the existing network topology model, perform new site selection planning and business configuration prediction; S4: Generate optimization instructions based on the planning results, and control the network management system to execute new site deployments and service adjustments.

2. The method for planning and analyzing new site services based on existing network topology according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain the geographical location information, cell parameters, and neighbor cell relationships of existing base stations from the network management system; S102: Collect live network service data, including traffic statistics, user distribution, and performance metrics. S103: Clean, correlate, and standardize the acquired data to obtain standardized live network topology data and business datasets.

3. The method for planning and analyzing new site services based on existing network topology according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Extract features from the standardized network topology data and service data to obtain topology features and service features; S202: Input the aforementioned features into the planning analysis model to construct a weighted topology graph model; S203: Identify areas with weak coverage and areas with high service load based on the topology model, and use these as areas for service planning requirements.

4. The method for planning and analyzing new site services based on existing network topology according to claim 1, characterized in that, Before executing step S2, it is necessary to train the parameters of the planning and analysis model based on historical business data and current network topology data; the planning and analysis model is used to comprehensively consider topology connectivity, interference relationships and business growth trends.

5. The method for planning and analyzing new site services based on existing network topology according to claim 1, characterized in that, The specific steps of S3 are as follows: S301: Determine a set of candidate site locations within the business planning requirement area; S302: Utilize optimization algorithms combined with the existing network topology model to select the optimal location and service configuration parameters for new sites; S303: Predicts the service offloading effect and network performance improvement indicators after the deployment of new sites.

6. The method for planning and analyzing new site services based on existing network topology according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Convert the planning results into new site deployment instructions, business parameter configuration instructions, and resource adjustment instructions; S402: Use the optimization instructions to drive the network management system to perform real-time deployment of new sites and service adjustments; S403: During execution, the existing network topology data is continuously fed back, the input of the planning analysis model is dynamically updated, and the optimization scheme is regenerated.

7. The method for planning and analyzing new site services based on existing network topology according to claim 2, characterized in that, The process described in S103 involves cleaning, associating, and standardizing the acquired data. Specifically, this includes spatially associating base station location information with geographic information system data, constructing an adjacency matrix based on neighboring cell relationships, and performing time and spatial aggregation processing on the business data to generate gridded business distribution data.

8. The method for planning and analyzing new site services based on existing network topology according to claim 3, characterized in that, The process described in S203, which identifies areas with weak coverage and high service load based on the topology graph model, involves calculating the topology centrality and service load index of each node, comparing the calculation results with preset coverage thresholds and capacity thresholds, and identifying areas that do not meet the thresholds as areas requiring service planning.

9. The method for planning and analyzing new site services based on existing network topology according to claim 5, characterized in that, The specific operation of S302 is as follows: construct a planning cost map that includes the current network topology model and the service planning requirement area; use heuristic search or linear programming methods to search for the optimal site deployment scheme on the planning cost map; perform service simulation verification on the selected deployment scheme to obtain the optimized service configuration parameters.

10. The method for planning and analyzing new site services based on existing network topology according to claim 6, characterized in that, The specific operation of S403 is to monitor the key performance indicators after the deployment of the new site in real time, and compare them with the predicted network performance improvement indicators; calculate the deviation between the actual service diversion effect and the predicted effect. When the deviation value exceeds the preset threshold, the planning analysis model is automatically triggered to correct the parameters and re-execute the new site selection planning or business configuration adjustment.

Images