Base station deployment planning method, device, equipment and program product

Abstract

The invention discloses a base station deployment planning method and device, equipment and a program product. According to the scheme, firstly, a target area candidate base station position set containing a plurality of candidate geographic positions is obtained and serves as an initial solution of a base station deployment scheme; then, on the basis of a Berfh triangle optimizer algorithm, iteration is carried out by taking optimization of a specified performance index as a target, in each round of iteration, for each candidate position, according to a relative relation between the candidate position and other positions in a current solution set, a gravitation direction of a position with better pointing performance is calculated, and whether the position is in a signal limited area is judged; for the position of the non-limited area, the position is updated along the gravitation direction by adopting a first type of adjustment so as to strengthen local development, and for the position in the limited area, a random jump component is introduced by adopting a second type of adjustment so as to promote the position to jump out of the current communication black hole area so as to strengthen global exploration; and finally, outputting a base station deployment scheme after iteration optimization.

Description

Technical Field

[0001] This application relates to the field of wireless network planning technology, and in particular to a base station deployment planning method, apparatus, equipment and program product. Background Technology

[0002] In the field of wireless network planning, traditional base station deployment methods generally suffer from insufficient dynamic response to environmental physical constraints. These methods typically rely on fixed rules or simple avoidance mechanisms to handle weak signal coverage areas or "communication black holes" (such as behind tall buildings or in complex terrain). However, this approach is often static and passive, unable to intelligently and differentiate the search strategy during optimization iterations. When candidate solutions get stuck in these signal-constrained areas, traditional algorithms are prone to converge to local optima due to the lack of effective escape mechanisms, resulting in coverage blind spots or uneven resource allocation in the final deployment scheme. Summary of the Invention

[0003] This application proposes a base station deployment planning method, apparatus, equipment, and program product, aiming to solve the key problems in existing technologies where optimization algorithms have poor adaptability to actual physical environments and are prone to getting trapped in local optima in complex geographical scenarios. Accordingly, the technical solution of this application is as follows: Firstly, a base station deployment planning method is provided, including: Obtain a set of candidate base station locations for the target area, wherein the set of candidate base station locations includes multiple candidate geographical locations for deployable base stations; Based on the Bermuda Triangle optimizer algorithm, the candidate base station location set is used as the initial solution for the base station deployment scheme in the target area. The base station deployment scheme is iterated with the goal of optimizing a specified performance index. Each iteration performs the following for each candidate location: Calculate the gravitational direction of the current candidate location based on its relative relationship with other candidate locations in the current solution set, where the gravitational direction points to the candidate location with the better specified performance index; determine whether the current candidate location is in a signal-restricted area; if not, perform a first type of adjustment, guiding the current candidate location to update its location according to the gravitational direction; if it is in the signal-restricted area, perform a second type of adjustment, guiding the current candidate location to update its location with the goal of escaping the signal-restricted area according to random jump components. Output the iterative base station deployment plan.

[0004] Secondly, a base station deployment planning method is provided, including: The preparation module is used to obtain a set of candidate base station locations in the target area, wherein the set of candidate base station locations contains multiple candidate geographical locations of deployable base stations; An optimization module is used to iterate on the base station deployment scheme based on the Bermuda Triangle optimizer algorithm, using the set of candidate base station locations as the initial solution for the base station deployment scheme in the target area, with the goal of optimizing a specified performance index of the base station deployment scheme. Each iteration performs the following for each candidate location: Calculate the gravitational direction of the current candidate location based on its relative relationship with other candidate locations in the current solution set, where the gravitational direction points to the candidate location with the better specified performance index; determine whether the current candidate location is in a signal-restricted area; if not, perform a first type of adjustment on the current candidate location, guiding the current candidate location to update its location according to the gravitational direction; if it is in the signal-restricted area, perform a second type of adjustment on the current candidate location, guiding the current candidate location to update its location with the goal of escaping the signal-restricted area according to random jump components. The output module is used to output the iterative base station deployment plan.

[0005] Thirdly, embodiments of this application provide an electronic device, including: a processor; and a memory configured to store computer-executable instructions, which, when executed, cause the processor to perform the method described in the first aspect.

[0006] Fourthly, a computer program product is provided, the computer program product including a computer-readable storage medium storing a computer program operable to cause a computer to perform the method described in the first aspect.

[0007] This application provides a base station deployment planning scheme aimed at solving the key problems of traditional optimization algorithms in complex geographical environments, which are prone to getting trapped in local optima due to a lack of physical perception capabilities and difficulty in dynamically avoiding signal-constrained areas (such as communication black holes). The scheme first obtains a set of candidate base station locations in a target area containing multiple candidate geographical locations, and uses this as the initial solution for the base station deployment scheme. Then, based on the Bermuda Triangle optimizer algorithm, it iterates with the goal of optimizing a specified performance index. In each iteration, for each candidate location, the gravitational direction pointing to a better-performing location is calculated based on its relative relationship with other locations in the current solution set, and it is determined whether the location is in a signal-constrained area. For locations in unconstrained areas, a first type of adjustment is used to update them along the gravitational direction to enhance local development; for locations in constrained areas, a second type of adjustment is used to introduce random jump components to encourage them to jump out of the current communication black hole region to enhance global exploration. Finally, the iteratively optimized base station deployment scheme is output. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a schematic diagram of the first process of the base station deployment planning method according to an embodiment of this application.

[0010] Figure 2 This is a schematic diagram of a second process for a base station deployment planning method according to an embodiment of this application.

[0011] Figure 3 This is a schematic diagram of the third process of the base station deployment planning method according to an embodiment of this application.

[0012] Figure 4 This is a schematic diagram of the fourth process of the base station deployment planning method according to an embodiment of this application.

[0013] Figure 5 This is a schematic diagram of the base station deployment planning device according to an embodiment of this application.

[0014] Figure 6 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0015] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0016] In the field of wireless network planning, traditional base station deployment methods suffer from insufficient dynamic response to environmental physical constraints. They typically rely on fixed rules or simple avoidance mechanisms to handle weak signal areas (such as "communication black holes" behind tall buildings or in complex terrain). These approaches are mostly static and passive, unable to differentiate and intelligently adjust search strategies during optimization iterations. This leads to candidate solutions converging to local optima once trapped in signal-constrained areas due to a lack of effective escape mechanisms, resulting in coverage blind spots or uneven resource allocation. Furthermore, most algorithms struggle to achieve an effective balance between global exploration and local development, especially in complex scenarios with numerous geographically restricted areas in the solution space. Existing solutions often employ single, uniform location update rules, failing to deeply integrate geographical constraints with iterative search strategies. This makes it difficult to coordinate coverage breadth, interference suppression, and construction costs, limiting the overall performance and automation level of network planning. Therefore, this application proposes a base station deployment planning method, apparatus, equipment, and program product, aiming to address the key problems of existing optimization algorithms' poor adaptability to actual physical environments and their tendency to get trapped in local optima in complex geographical scenarios. The technical solutions provided by the various embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0017] One embodiment of this application provides a base station deployment planning method. Figure 1 This is a flowchart illustrating the base station deployment planning method, including: S101, Obtain the candidate base station location set for the target area. The candidate base station location set contains multiple candidate geographical locations for deployable base stations.

[0018] In this embodiment, the "target area" refers to a specific geographical area where wireless network (such as a cellular mobile communication network) planning and base station deployment are required, such as a city, an industrial park, or a continuous area that needs coverage. This area typically has clear geographical boundaries and may contain diverse terrain, building distribution, and other physical environmental features that affect radio wave propagation. The "candidate base station location" refers to a geographical location within the target area that, after preliminary evaluation from an engineering construction and wireless coverage perspective, is considered potentially feasible. These locations are considered as alternative sites for the final base station deployment. Each candidate geographical location corresponds to a specific coordinate (such as latitude and longitude or planar coordinates) and represents a possible base station installation point.

[0019] The reason this embodiment requires obtaining a set of candidate base station locations in the target area is to provide a structured, non-empty initial search space for subsequent intelligent optimization algorithms. Traditional methods often conduct aimless searches in continuous geographic space, easily getting bogged down in areas that are completely infeasible due to physical or policy constraints, leading to low computational efficiency. Through this step, we transform the continuous solution space, which may be full of "no-go zones," into a discrete "solution pool" consisting of a finite number of initially screened feasible points. This not only significantly improves optimization efficiency, allowing the algorithm to focus on valuable candidate points for combined optimization, but also ensures that the final solution has stronger engineering feasibility, as the candidate point set itself incorporates basic engineering constraints. More importantly, this step lays the foundation for subsequent implementation of differentiated intelligent adjustment strategies: each candidate location in the set can be pre-labeled with attributes (e.g., whether it is in a "restricted area" where signal propagation is easily obstructed), enabling subsequent algorithms to dynamically bind different search strategies based on location attributes, thereby accurately adapting to actual physical environment constraints.

[0020] In its implementation, this embodiment first utilizes low-altitude drones, ground sensors, and other equipment to collect raw data such as terrain data, user density distribution, and existing base station information for the target area. This data is integrated and processed through a geographic information system (GIS) to mark prohibited "communication black hole" areas and generate a binary mask matrix to identify feasible and forbidden zones. Subsequently, the collected data undergoes denoising, normalization, and gridding preprocessing to eliminate outliers and standardize dimensions. Finally, a set of candidate base station locations is formed through spatial grid partitioning. This process ensures that the candidate locations both cover the hotspot needs of the target area and comply with engineering constraints, providing high-quality input for subsequent algorithm iterations.

[0021] S102, based on the Bermuda Triangle optimizer algorithm, uses the set of candidate base station locations as the initial solution for the base station deployment scheme in the target area. It iterates through the base station deployment scheme with the goal of optimizing a specified performance index. Each iteration performs the following for each candidate location: Calculate the gravitational direction of the current candidate location based on its relative relationship with other candidate locations in the current solution set, where the gravitational direction points to the candidate location with the better specified performance index; determine whether the current candidate location is in a signal-restricted area; if not, perform a first-type adjustment, guiding the current candidate location to update its position according to the gravitational direction; if it is in a signal-restricted area, perform a second-type adjustment, guiding the current candidate location to update its position with the goal of escaping the signal-restricted area according to random jump components.

[0022] This embodiment employs the Bermuda Triangle Optimizer algorithm to iteratively optimize base station deployment schemes. This algorithm is a heuristic optimization method inspired by the Bermuda Triangle phenomenon, exploring the solution space by simulating gravitational fields and anomalous movement mechanisms. Using the set of candidate base station locations as the initial solution for the base station deployment scheme in the target area is significant in providing diverse starting points for the algorithm, preventing the deployment scheme from getting trapped in local optima, and thus systematically optimizing specified performance indicators such as signal coverage, signal interference levels, base station energy consumption, and scenario type adaptability. Iterative optimization based on the Bermuda Triangle Optimizer algorithm effectively integrates the physical propagation characteristics of 5G signals, improving the robustness and accuracy of the deployment scheme in complex environments and reducing reliance on manual parameter tuning.

[0023] In each iteration, the core operations performed for each candidate position include calculating the gravitational direction, identifying signal-constrained regions, and performing differentiated adjustments. When calculating the gravitational direction, based on the relative relationship between the current candidate position and other candidate positions in the current solution set, the gravitational direction is directed towards the candidate position with better performance indicators. This guides the solution to move towards a better region, gradually optimizing the overall deployment. Determining whether the current candidate position is in a signal-constrained region (i.e., a "communication black hole" region, such as a tall building obstruction area or a terrain-restricted area) is a crucial step. If it is not in a signal-constrained region, a first type of adjustment is performed, smoothly updating the position according to the gravitational direction to ensure steady optimization within a safe area. If it is in a signal-constrained region, a second type of adjustment is triggered, forcibly escaping local traps according to random jump components (such as random vectors following a Lévy distribution), simulating abnormal movement to avoid obstruction. This dual-mechanism combination ensures the algorithm's adaptability in dynamic environments, enabling both fine-grained development of feasible regions and global exploration to avoid dead zones.

[0024] In calculating the direction of gravity, this embodiment constructs a gravitational field model simulating the signal coverage of the target area. The gravitational field strength at any target location is determined by its relative relationship to at least one candidate location with better specified performance indicators. The calculation of this relative relationship deeply integrates path loss and terrain occlusion parameters of the communication signal, allowing the gravity value to directly reflect the propagation quality of the 5G signal. For example, the gravitational field model can be expressed by the formula:

[0025] in, Indicates candidate position With the center position The gravitational field between them is strong; It is the initial gravitational constant, corresponding to the baseline propagation capability of 5G signals (using 5G signals as an example only), and its value is related to the frequency band; It is the quality of the central solution, which is dynamically updated to the best covered solution in the current population; The quality of the candidate solution is defined as the effective coverage area per unit energy consumption. It is the distance between the candidate position and the center position; It is the distance attenuation factor, set according to the signal propagation model; It is the terrain height difference factor, calculated using a digital elevation model; It refers to the altitude difference. By mapping the current candidate location to this gravitational field model, the direction of gravity can be quantitatively determined, aligning the optimization process with physical laws.

[0026] Furthermore, this embodiment integrates a reinforcement learning model to enhance dynamic adaptability. Real-time network status data of the target area is collected, including base station location, transmit power, number of users, signal interference matrix, and geographical no-go zones. This data drives the reinforcement learning model to make adjustments to the current base station deployment plan, such as adjusting candidate base station locations, base station transmit power, or base station frequency bands. The decision strategy of the reinforcement learning model is optimized through a reward function, which is coupled with a specified performance indicator. For example, the reward function can be designed as a weighted sum of changes in coverage efficiency, interference index, and energy consumption, thereby ensuring that the action selection is geared towards overall performance improvement. Based on the model's output adjustment actions, the current base station deployment plan is updated in real time, forming a closed-loop optimization of "perception-decision-execution".

[0027] After obtaining the iterative base station deployment plan, this embodiment further refines the candidate base station locations based on the edge computing requirements of the target area and the adaptability requirements of different scenario types. Scenario types include urban scenarios, mountainous scenarios, vehicle-to-everything (V2X) scenarios, IoT scenarios, and emergency communication scenarios. For example, interference suppression is emphasized in densely populated urban areas, while energy consumption optimization is prioritized in mountainous areas, achieving differentiated deployment through dynamic parameter configuration. This step ensures that the plan is not only theoretically optimal but also conforms to actual engineering constraints and business needs, improving the feasibility and efficiency of deployment. Figure 2This paper presents a closed-loop process for dynamic optimization of base station deployment driven by edge computing nodes. The process begins with real-time monitoring of network operation. When a pre-defined trigger event is detected—for example, a sudden change in user density in a hotspot area in an urban scenario, a base station issuing an alarm due to severe weather in a mountainous scenario, a significant change in traffic flow predicted in a vehicle-to-everything (V2X) scenario, or a disaster warning received in an emergency communication scenario—the system immediately activates a differentiated intelligent response mechanism. This mechanism deeply integrates scenario characteristics: for changes in user density, a lightweight Bermuda Triangle optimizer algorithm is run, and the algorithm weights are dynamically adjusted according to scenario objectives (e.g., interference suppression in urban areas, energy consumption optimization in mountainous areas) to quickly optimize local base station parameters; for base station failures, backup nodes are activated or coverage is reconstructed using wireless self-backhaul technology, based on the scenario's network architecture (e.g., relay links in mountainous areas, mesh networking in emergency scenarios); for external disturbances such as weather, power is dynamically increased or the system switches to redundant links based on a link reliability model. After each optimization action, the system updates the local resource allocation status in real time and securely uploads the anonymized key optimization data to the cloud via a federated learning framework. The cloud then aggregates the experience from multiple edge nodes to update the global optimization model, thus achieving a continuous adaptive closed loop of "scene awareness - event triggering - differentiated optimization - federated feedback." This mechanism ensures that the final deployment solution not only converges to the global optimum in theory but also deeply adapts to dynamically changing real-world engineering environments and diverse business scenarios, significantly improving the resilience, efficiency, and intelligence of network deployment.

[0028] S103 outputs the iterative base station deployment scheme.

[0029] After completing the iterative optimization based on the Bermuda Triangle optimizer algorithm, this embodiment does not directly output the algorithm-generated solution. Instead, it first performs high-fidelity engineering verification to ensure its effectiveness and reliability in a real-world environment. This verification process is achieved through digital twin technology: according to the iterative base station deployment scheme, a digital twin, i.e., a virtual communication environment, is constructed in digital space that is highly consistent with the physical environment, electromagnetic characteristics, and service requirements of the target area. This virtual environment integrates accurate geographic information system data, 3D building models, measured or simulated wireless channel models, and network equipment parameters. In this environment, the iterative deployment scheme can be comprehensively simulated and verified to evaluate its coverage performance, interference level, capacity carrying capacity, and other key performance indicators in complex real-world scenarios. Only when the simulation verification results meet the preset performance thresholds and engineering constraints is the scheme confirmed as "verified successfully." This step essentially places the "theoretically optimal" scheme generated by the intelligent algorithm in a near-realistic virtual environment for stress testing and feasibility confirmation, thereby effectively avoiding the scheme defects caused by model simplification or incomplete assumptions, and significantly improving the engineering feasibility and success rate of the output scheme. Only after passing this verification will the iterative base station deployment plan be finally output.

[0030] The output base station deployment plan is an engineering planning outcome with direct guiding significance. Its main uses are reflected in the following aspects: First, it serves as a direct technical basis for network construction and implementation. The geographical location and configuration recommendations for each candidate base station specified in the plan (which can be combined with the adjustment actions output by the reinforcement learning model in step S102, such as the recommended transmit power or frequency band) constitute the foundation for construction drawing design and equipment procurement. Second, the plan can serve as a benchmark reference for network optimization and dynamic adjustment, providing an authoritative initial state and optimization starting point for subsequent network operation and maintenance, capacity expansion, and dynamic reconfiguration based on real-time status. Finally, the plan is of great value for network investment decisions and benefit assessment. It quantitatively predicts the network performance level achievable under a specific investment (number and location of base stations), providing objective, data-driven decision support for planners to compare the cost-effectiveness of different construction strategies. Therefore, the output is not just a set of location coordinates, but a comprehensive, executable, and evaluable network planning blueprint that integrates intelligent algorithm optimization, digital environment verification, and is closely coupled with business objectives.

[0031] In summary, the method of this embodiment first obtains a set of candidate base station locations in a target area containing multiple candidate geographical locations, and uses this as the initial solution for the base station deployment scheme. Then, based on the Bermuda Triangle optimizer algorithm, it iterates with the goal of optimizing a specified performance index. In each iteration, for each candidate location, the gravitational direction pointing to the location with better performance is calculated according to its relative relationship with other locations in the current solution set, and it is determined whether the location is in a signal-restricted area. For locations in unrestricted areas, a first type of adjustment is used to update them along the gravitational direction to enhance local development, while for locations in restricted areas, a second type of adjustment is used to introduce random jump components to make them jump out of the current communication black hole region to enhance global exploration. Finally, the iteratively optimized base station deployment scheme is output.

[0032] The process of the base station deployment planning method in this embodiment will be described in detail below.

[0033] refer to Figure 3 As shown, the base station deployment planning method in this embodiment is based on the Bermuda Triangle optimizer algorithm, aiming to solve the multi-objective coordination problem of coverage, interference, and energy consumption in base station deployment through intelligent iterative optimization. The process begins with multi-source data acquisition and mainly includes the following data types and corresponding processing procedures: For terrain data, a digital elevation model (DEM) is imported to generate a digital elevation network describing the terrain undulations. Then, geographical features (such as high-rise buildings and complex terrain) are identified and marked, and finally, a binary mask matrix M that identifies prohibited deployment areas is output, providing clear geospatial constraints for subsequent optimization.

[0034] For business data, user density analysis and heatmap creation, combined with traffic prediction models, dynamically depict the distribution and changing trends of business demand in the target area, forming a standardized input reflecting the spatial characteristics of business load. For existing base station data, information including base station location, transmit power, and historical interference records is collected and unified to the same spatial grid reference as the terrain and business data through network coordinate matching. After feature extraction and normalization, the above-mentioned data types enter the data fusion stage, where coordinate alignment, format unification, and normalization are performed, ultimately outputting a standardized dataset integrating geographical constraints, business distribution, and network status. This dataset serves as a high-quality structured input for the subsequent Bermuda Triangle optimizer algorithm, laying a reliable data foundation for the intelligent iterative optimization of base station deployment schemes.

[0035] In practical implementation, the first step is to construct a candidate solution space driven by multi-source data. Low-altitude drones, ground sensors, and other equipment are used to collect multi-dimensional data on the target area, including terrain elevation, building distribution, user density heatmaps, and existing base station information. This raw data is then integrated and processed by a geographic information system (GIS). A key step is to identify and mark "communication black hole" areas formed by terrain occlusion or building clusters, and to generate a binary geographic no-go zone mask matrix. , where matrix elements The location of the marker coordinates is within a physically restricted area where deployment is prohibited. The data then undergoes denoising, normalization, and spatial gridding preprocessing to ultimately form a set of candidate base station locations covering the target area and preliminarily avoiding core restricted zones. This set not only provides a high-quality discrete search space for subsequent optimization, but its attached "restricted" attribute label also forms the basis for decision-making in implementing differentiated optimization strategies.

[0036] After obtaining the set of candidate base station locations, this embodiment constructs a multi-objective optimization model. Its objective function comprehensively considers coverage efficiency, interference index, energy consumption cost, and scenario adaptability. The formula is as follows:

[0037] in, Represents coverage efficiency, and calculates the effective coverage area based on Shannon's formula and ray tracing algorithm; The interference index is used to assess the intensity of co-channel interference through a signal power superposition model. It represents the energy consumption function and models the nonlinear relationship between base station power and operating power consumption; This represents the security margin adapted to specific scenarios, quantifying the degree of avoidance of "communication black hole" regions. Weighting coefficients. , ... and It supports dynamic configuration to adapt to different business needs, such as improving efficiency in emergency communication scenarios. With a focus on coverage, improve in green and energy-saving scenarios, The model emphasizes energy consumption. It also includes multi-dimensional constraints such as geospatial constraints, resource constraints, and business quality constraints. Ensure candidate positions Avoid the mask matrix Marked restricted areas; power constraints Ensure transmit power vector Within reasonable limits; spectrum constraints The set of adjacent base station frequency bands is obtained using a graph coloring algorithm. and Orthogonal allocation to avoid interference; business quality constraints , This ensures the number of users per base station. Not exceeding the capacity limit And signal-to-interference-plus-noise ratio Not lower than the threshold .

[0038] The iterative optimization process is the core step. Each iteration executes a dual-region update strategy for each candidate position. First, based on the relative relationship between the current candidate position and other positions in the current solution set, the gravitational direction is calculated. This direction points to the candidate position with the better performance index. The gravitational field model is implemented through physical layer mapping, and the formula, as described above, can be expressed as:

[0039] Based on the above physical model, this embodiment implements a dual-region solution update mechanism. In each iteration, the algorithm updates the solution based on the geographical restricted area mask matrix. Determine whether each candidate location is in a communication black hole.

[0040] For those not in signal-restricted areas ( The candidate positions are determined using a first-type adjustment strategy, namely, gravity-driven smooth movement, and the position update formula is as follows:

[0041] in, Candidate position In the The coordinates at the next iteration; This is the gravitational response coefficient; It is the gravitational gradient vector; It is the inertial velocity vector; It is a dynamic inertia weight, the value of which decays with iteration, and the calculation formula is: This is to achieve a smooth transition from global search to local development. and These are the upper and lower limits of the inertia weight, respectively; Control weights from decay to near The speed. This design allows the algorithm to perform at a high speed in the early stages of iteration ( When the size is large, it relies more on inertial exploration to enhance global search capabilities; in the later stages of iteration ( When the gravity gradient is smaller, it tends to make fine adjustments following the gravity gradient, thus achieving a smooth transition from global exploration to local development.

[0042] For areas with limited signal ( If the candidate position is not found, the second type of adjustment strategy is forcibly triggered, employing Levi flight to perform abnormal movement to escape the local trap. The position update formula is:

[0043] in, It is a coordinate randomly selected from the safe area; It is the scaling factor; It is a random vector that follows a Lévy distribution, and its probability density function is: The Lévy distribution has a heavy-tailed characteristic, which can generate long-step jumps, effectively simulating anomalous movement to avoid terrain occlusion.

[0044] Meanwhile, to cope with the dynamically changing network environment, this implementation method introduces a reinforcement learning closed-loop optimization system. This system defines the state. Including time Base station location vector Power vector User number vector Interference matrix and geographical restricted area mask The action space is defined as follows: It supports adjustments to location, power, and frequency band, or triggering re-optimization. The reward function is designed as follows: ,in , and These represent the changes in coverage efficiency, interference index, and energy consumption relative to the previous time step, respectively. The strategy is trained using a deep Q-network algorithm, and its loss function is... , , As a discount factor, and The parameters are set for the online network and the target network, respectively, thus forming a real-time optimization closed loop of "perception-decision-execution".

[0045] Building upon the above, this embodiment designs a multi-scenario adaptive deployment strategy for diverse practical application scenarios. In densely populated urban areas, a three-dimensional ray tracing model is introduced to optimize the antenna tilt angle. And a frequency band matrix is ​​allocated using a fractional frequency reuse strategy. To suppress interference. In remote mountainous areas, a hybrid architecture of "macro base station + relay station" is adopted, and the coverage efficiency of the relay station is maximized. To optimize relay station locations ,in For user coordinates, User priority weights are assigned. In vehicle-to-everything (V2X) scenarios, LSTM neural networks are used to predict vehicle trajectories and guide the pre-deployment of micro base stations to critical path nodes. In IoT scenarios, access capacity constraints are established. And implement a device hibernation grouping strategy. To reduce energy consumption. In emergency communication scenarios, a priority function for temporary base station site selection is established based on disaster heat maps. In this function, This indicates the emergency coverage priority score for candidate locations; Estimated number of users in the disaster-stricken area surrounding the candidate site; It is a level coefficient (e.g., 1-10) that characterizes the severity of a disaster. and These represent the geometric distances of the candidate points to the nearest hospital and fire station, respectively. By maximizing this function value through an optimization algorithm, the locations of base stations that prioritize coverage of critical areas such as medical facilities and rescue operations can be determined. Furthermore, to improve coverage flexibility, vehicle-mounted or airborne mobile base stations are used for dynamic scheduling, with their location updates following a formula. To achieve dynamic coverage, among which, It is the mobile base station at any time Location coordinates, It is its movement speed. It is the time interval for scheduling updates. It is an emergency priority function The gradient at the current base station location indicates the direction of movement toward a higher priority area.

[0046] In addition, this embodiment also uses digital twin technology to construct a high-fidelity virtual environment to pre-verify the deployment scheme. The verification process follows:

[0047] Simultaneously, AI-generated content technology is used to expand the diversity of training data. Local real-time optimization is achieved through edge computing architecture, and a federated learning formula is employed. Update global model parameters ,in For the first Local model parameters for each edge node Its sample size.

[0048] In addition, refer to Figure 4 As shown, this embodiment also establishes a dynamic resource collaborative optimization framework to jointly schedule base station location, transmit power, spectrum resources, computing resources, and transmission resources. This framework unifies the multi-dimensional resources into a state space model. .in, Represents the location coordinate vector of all base stations; The transmit power vector representing the base station; This represents the vector of frequency band sets allocated to each base station; This represents a vector of edge computing resources, whose elements include the number of CPU cores, GPU computing power, memory capacity, etc. This represents a transmission resource vector, whose elements include fiber optic link bandwidth, wireless backhaul latency, etc.

[0049] The collaborative optimization objective under this framework is: .in, This represents the set of multidimensional resource states to be optimized. to These are the weighting coefficients for each indicator; , and These are the coverage efficiency, interference index, and energy consumption function mentioned above, respectively. This indicates the utilization rate of computing resources, used to avoid over-allocation or idleness of computing resources; This indicates end-to-end communication latency, designed to ensure the service quality of real-time services such as autonomous driving.

[0050] To achieve efficient spectrum resource utilization, a graph-based spectrum coloring algorithm is used to dynamically allocate frequency bands. The allocation rules are as follows: In this formula, Indicates to be assigned to base station Frequency band index; Index of available frequency bands; A set of indices representing all available frequency bands; Indicates connection with base station A set of neighboring base stations that pose potential interference. This rule ensures that adjacent base stations are not assigned the same frequency band.

[0051] In addition, to intelligently schedule edge computing resources, this embodiment also designs a task offloading decision model. .in, Indicates the task The uninstallation decision (local processing or uninstallation to the cloud); It is a task The computational requirements; It is the upper limit of computing power for edge nodes; k is the task Maximum allowable processing latency; It is the estimated latency for task processing at the edge node.

[0052] At the same time, through the computing power elastic allocation formula This enables on-demand resource scaling. In this formula, e represents the number of virtual machines that need to be dynamically adjusted; This represents the system's current total computing load; This represents the standard computing power unit that a single virtual machine instance can provide (such as 1 vCPU and 2GB of memory).

[0053] Finally, the Pareto optimal front under this framework is solved using multi-objective optimization algorithms such as NSGA-II, thereby providing network operators with a resource allocation scheme that achieves the best trade-offs in multiple dimensions such as coverage quality, energy consumption, and service latency.

[0054] In addition, corresponding to Figure 1 In addition to the method shown in this embodiment, another embodiment of this invention also provides a base station deployment planning device. Figure 5 This is a structural diagram of the base station deployment planning device 500, including: The preparation module 510 is used to obtain a set of candidate base station locations in the target area, wherein the set of candidate base station locations includes multiple candidate geographical locations of deployable base stations.

[0055] The optimization module 520 is used to iterate the base station deployment scheme based on the Bermuda Triangle optimizer algorithm, using the candidate base station location set as the initial solution for the base station deployment scheme in the target area, with the goal of optimizing a specified performance index of the base station deployment scheme. Each iteration performs the following for each candidate location: Calculate the gravitational direction of the current candidate location based on its relative relationship with other candidate locations in the current solution set, where the gravitational direction points to the candidate location with the better specified performance index; determine whether the current candidate location is in a signal-restricted area; if not, perform a first type of adjustment on the current candidate location, guiding the current candidate location to update its location according to the gravitational direction; if it is in the signal-restricted area, perform a second type of adjustment on the current candidate location, guiding the current candidate location to update its location with the goal of escaping the signal-restricted area according to a random jump component.

[0056] Output module 530 is used to output the iterative base station deployment scheme.

[0057] Optionally, the optimization module 520 calculates the gravitational direction of the current candidate position based on the relative relationship between the current candidate position and other candidate positions in the current solution set, including: constructing a gravitational field model simulating the signal coverage of the target area; wherein, the gravitational field strength of any target position in the gravitational field model is determined by the relative relationship between the target position and at least one candidate position with better specified performance index, and the calculation of the relative relationship incorporates the path loss of communication signals and terrain occlusion parameters; mapping the current candidate position to the gravitational field model to determine the gravitational direction of the current candidate position.

[0058] Optionally, the optimization module 520 is further configured to: collect network status data of the target area in real time, the network status data including at least one of base station location, transmission power, number of users, signal interference matrix, and geographical no-go zone; based on the network status data, determine adjustment actions for the current base station deployment scheme through a reinforcement learning model, the adjustment actions including at least one of adjusting candidate base station location, adjusting base station transmission power, and adjusting base station frequency band; wherein, the decision strategy of the reinforcement learning model is optimized through a reward function, the reward function being coupled with the specified performance index; and update the current base station deployment scheme based on the adjustment actions determined by the reinforcement learning model.

[0059] Optionally, the optimization module 520 is further configured to: before outputting the iterative base station deployment scheme, adjust the candidate base station positions in the iterative base station deployment scheme based on the edge computing requirements related to the target area and / or the adaptability requirements of the scenario type of the target area; wherein, the scenario type includes at least one of urban scenario, mountainous scenario, vehicle-to-everything (V2X) scenario, Internet of Things (IoT) scenario, and emergency communication scenario.

[0060] Optionally, the optimization module 520 is further configured to: before outputting the iterative base station deployment scheme, construct a virtual communication environment for the target area based on digital twin technology and in accordance with the iterative base station deployment scheme, and verify the iterative base station deployment scheme based on the virtual communication environment.

[0061] Optionally, the random jump component used in the second type of adjustment follows a Lévy distribution.

[0062] Optionally, the specified performance indicators include at least one of signal coverage effect, signal interference level, base station energy consumption, and scene type adaptability.

[0063] It should be noted that the base station deployment planning device in this embodiment can be used as... Figure 1 The execution body of the method shown is therefore able to achieve... Figure 1 The steps and functions of the method shown are illustrated.

[0064] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Please refer to it. Figure 6 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and memory. The memory may include main memory, such as high-speed random-access memory (RAM), or non-volatile memory, such as at least one disk drive. Of course, the electronic device may also include other hardware required for other business operations.

[0065] The processor, network interface, and memory can be interconnected via an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0066] Memory is used to store computer programs. Specifically, a computer program may include program code, which includes computer operation instructions. Memory may include main memory and non-volatile memory, and provides the computer program to the processor.

[0067] Specifically, the processor reads the corresponding computer program from non-volatile memory into memory and then runs it, forming the above-mentioned logical structure. Figure 5 The base station deployment planning device shown. Correspondingly, the processor executes the program stored in the memory, and specifically performs the following operations: Obtain a set of candidate base station locations for the target area, wherein the set of candidate base station locations contains multiple candidate geographical locations for deployable base stations.

[0068] Based on the Bermuda Triangle optimizer algorithm, the candidate base station location set is used as the initial solution for the base station deployment scheme in the target area. The base station deployment scheme is iterated with the goal of optimizing a specified performance index. Each iteration performs the following for each candidate location: Calculate the gravitational direction of the current candidate location based on its relative relationship with other candidate locations in the current solution set, where the gravitational direction points to the candidate location with the better specified performance index; determine whether the current candidate location is in a signal-restricted area; if not, perform a first type of adjustment, guiding the current candidate location to update its position according to the gravitational direction; if it is in the signal-restricted area, perform a second type of adjustment, guiding the current candidate location to update its position with the goal of escaping the signal-restricted area according to random jump components.

[0069] Output the iterative base station deployment plan.

[0070] The above is as described in this instruction manual. Figure 1The base station deployment planning method disclosed in the illustrated embodiments can be applied to a processor and implemented by the processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit in the processor or by instructions in the form of software. The processor mentioned above can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this application can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0071] Of course, in addition to software implementation, the electronic device described in this specification does not exclude other implementation methods, such as logic devices or a combination of hardware and software. In other words, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0072] Furthermore, embodiments of this application also propose a computer program product, including a computer-readable storage medium storing one or more computer programs, the one or more computer programs including instructions.

[0073] When the aforementioned instructions are executed by a portable electronic device that includes multiple applications, they enable the portable electronic device to perform... Figure 1 The steps in the method shown include: Obtain a set of candidate base station locations for the target area, wherein the set of candidate base station locations contains multiple candidate geographical locations for deployable base stations.

[0074] Based on the Bermuda Triangle optimizer algorithm, the candidate base station location set is used as the initial solution for the base station deployment scheme in the target area. The base station deployment scheme is iterated with the goal of optimizing a specified performance index. Each iteration performs the following for each candidate location: Calculate the gravitational direction of the current candidate location based on its relative relationship with other candidate locations in the current solution set, where the gravitational direction points to the candidate location with the better specified performance index; determine whether the current candidate location is in a signal-restricted area; if not, perform a first type of adjustment, guiding the current candidate location to update its position according to the gravitational direction; if it is in the signal-restricted area, perform a second type of adjustment, guiding the current candidate location to update its position with the goal of escaping the signal-restricted area according to random jump components.

[0075] Output the iterative base station deployment plan.

[0076] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0077] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0078] The above are merely embodiments of this specification and are not intended to limit the scope of this specification. Various modifications and variations can be made to this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of the claims of this specification. Furthermore, all other embodiments obtained by those skilled in the art without inventive effort should fall within the protection scope of this document.

Claims

1. A base station deployment planning method, characterized in that, include: Obtain a set of candidate base station locations for the target area, wherein the set of candidate base station locations includes multiple candidate geographical locations for deployable base stations; Based on the Bermuda Triangle optimizer algorithm, the candidate base station location set is used as the initial solution for the base station deployment scheme in the target area. The base station deployment scheme is iterated with the goal of optimizing a specified performance index. Each iteration performs the following for each candidate location: Calculate the gravitational direction of the current candidate location based on its relative relationship with other candidate locations in the current solution set, where the gravitational direction points to the candidate location with the better specified performance index; determine whether the current candidate location is in a signal-restricted area; if not, perform a first type of adjustment, guiding the current candidate location to update its location according to the gravitational direction; if it is in the signal-restricted area, perform a second type of adjustment, guiding the current candidate location to update its location with the goal of escaping the signal-restricted area according to random jump components. Output the iterative base station deployment plan.

2. The method according to claim 1, characterized in that, Based on the relative relationship between the current candidate position and other candidate positions in the current solution set, the gravitational direction of the current candidate position is calculated, including: Construct a gravitational field model to simulate the signal coverage of the target area; wherein, the gravitational field strength at any target location in the gravitational field model is determined by the relative relationship between the target location and at least one candidate location with better specified performance index, and the calculation of the relative relationship incorporates the path loss of communication signals and terrain occlusion parameters; The current candidate position is mapped onto the gravitational field model to determine the gravitational direction of the current candidate position.

3. The method according to claim 1, characterized in that, Also includes: Real-time collection of network status data in the target area, including at least one of base station location, transmission power, number of users, signal interference matrix, and geographical no-go zone; Based on the network state data, a reinforcement learning model is used to determine adjustment actions for the current base station deployment scheme. The adjustment actions include at least one of adjusting the candidate base station location, adjusting the base station transmit power, and adjusting the base station frequency band. The decision strategy of the reinforcement learning model is optimized through a reward function, which is coupled with the specified performance index. The current base station deployment plan is updated based on the adjustment actions determined by the reinforcement learning model.

4. The method according to claim 1, characterized in that, Before outputting the iterative base station deployment scheme, the method includes: Based on the edge computing requirements related to the target area and / or the adaptability requirements of the target area's scenario type, the candidate base station locations in the iterative base station deployment scheme are adjusted; wherein, the scenario type includes at least one of urban scenario, mountainous scenario, vehicle-to-everything (V2X) scenario, Internet of Things (IoT) scenario, and emergency communication scenario.

5. The method according to claim 4, characterized in that, Before outputting the iterative base station deployment scheme, the method further includes: Based on digital twin technology, a virtual communication environment for the target area is constructed according to the iterative base station deployment scheme, and the iterative base station deployment scheme is verified based on the virtual communication environment.

6. The method according to any one of claims 1 to 5, characterized in that, The random jump component used in the second type of adjustment follows a Lévy distribution.

7. The method according to any one of claims 1 to 5, characterized in that, The specified performance indicators include at least one of the following: signal coverage effect, signal interference level, base station energy consumption, and scene type adaptability.

8. A base station deployment planning device, characterized in that, include: The preparation module is used to obtain a set of candidate base station locations in the target area, wherein the set of candidate base station locations contains multiple candidate geographical locations of deployable base stations; An optimization module is used to iterate on the base station deployment scheme based on the Bermuda Triangle optimizer algorithm, using the set of candidate base station locations as the initial solution for the base station deployment scheme in the target area, with the goal of optimizing a specified performance index of the base station deployment scheme. Each iteration performs the following for each candidate location: Calculate the gravitational direction of the current candidate location based on its relative relationship with other candidate locations in the current solution set, where the gravitational direction points to the candidate location with the better specified performance index; determine whether the current candidate location is in a signal-restricted area; if not, perform a first type of adjustment on the current candidate location, guiding the current candidate location to update its location according to the gravitational direction; if it is in the signal-restricted area, perform a second type of adjustment on the current candidate location, guiding the current candidate location to update its location with the goal of escaping the signal-restricted area according to random jump components. The output module is used to output the iterative base station deployment plan.

9. An electronic device, comprising: processor; And a memory arranged to store computer-executable instructions, characterized in that, when executed, the executable instructions cause the processor to perform the method as described in any one of claims 1 to 7.

10. A computer program product, the computer program product comprising a computer-readable storage medium storing a computer program, characterized in that, The computer program is operable to cause the computer to perform the method as described in any one of claims 1 to 7.

Images