An air-ground cooperative unmanned aerial vehicle efficiency perception virtual force dynamic deployment method and system

Abstract

The application discloses an air-ground cooperative unmanned aerial vehicle efficiency sensing virtual force dynamic deployment method and system, relates to the technical field of wireless communication networks, and comprises the following steps: acquiring current network state information, calculating based on the current network state information, obtaining user QoS deficit weight and adaptive capability step length, generating user attraction based on the user QoS deficit weight, performing kinematic damping correction calculation based on the user attraction and the adaptive capability step length, and obtaining corrected three-dimensional displacement; updating the unmanned aerial vehicle position based on the corrected three-dimensional displacement, updating energy consumption and service state based on the updated unmanned aerial vehicle position, and realizing dynamic deployment of the unmanned aerial vehicle.

Description

Technical Field

[0001] This invention relates to the field of wireless communication network technology, specifically to a method and system for dynamic deployment of virtual forces for energy efficiency perception of air-ground collaborative UAVs. Background Technology

[0002] To address the spatial deployment problem in user-centric, non-cellular, large-scale MIMO scenarios assisted by drones, existing technologies typically employ two types of approaches. One type is mathematical optimization methods, such as continuous convex approximation, alternating optimization, and block optimization. These methods usually jointly model drone position, user association, power control, and trajectory variables, and obtain an approximate optimal solution through iterative solving. The other type is heuristic or metaheuristic intelligent optimization methods, such as genetic algorithms, particle swarm optimization, gray wolf optimization, sparrow search algorithms, and their multi-objective extensions. These methods obtain deployment schemes through population initialization, fitness evaluation, and multiple rounds of search.

[0003] Existing technologies lack a dynamic deployment process in air-ground collaborative non-cellular massive MIMO networks that simultaneously considers user fairness, communication performance, UAV load balancing, real-time online adjustment, trajectory smoothing, safety constraints, and residual energy perception. This makes it impossible to achieve low-complexity, real-time dynamic deployment of UAVs that jointly perceives user fairness and residual energy. Summary of the Invention

[0004] To address the shortcomings mentioned in the background section, the present invention aims to provide a method and system for dynamic deployment of virtual forces for energy efficiency perception of air-ground collaborative unmanned aerial vehicles (UAVs).

[0005] Firstly, the objective of this invention can be achieved through the following technical solution: a method for dynamically deploying virtual forces for energy efficiency perception of air-ground cooperative UAVs, the method comprising the following steps: Obtain current network status information, which includes: the horizontal position coordinates of the user equipment, the current downlink achievable rate of the user, the location of the ground access point, the three-dimensional location of the UAV air access point, the large-scale channel gain between the user and each access point, the user's current service cluster, the remaining battery power of the UAV, the speed or displacement status of the UAV at the previous moment, and the set of users currently served by the UAV. The user QoS deficit weight and adaptive capability step size are calculated based on the current network state information. The user attraction is generated based on the user QoS deficit weight. The kinematic damping correction is calculated based on the user attraction and adaptive capability step size to obtain the corrected three-dimensional displacement. The drone's position is updated based on the corrected 3D displacement, and the energy consumption and service status are updated based on the updated drone position, enabling dynamic deployment of the drone.

[0006] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the calculation process of the user QoS deficit weight, as follows: Based on the user's current downlink achievable rate within the current network state information, and using the inverse rate normalization method, the user's QoS deficit weight is calculated. In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the magnitude of the user QoS deficit weight is inversely proportional to the magnitude of the user's current downlink achievable rate.

[0007] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the calculation process of the adaptive capability step size, as follows: The energy scaling factor is calculated based on the drone's remaining battery power and a preset safe battery power threshold, according to the current network status information. A feedback adjustment factor is generated based on the user's current downlink achievable rate according to the current network status information. Based on the energy scaling factor, feedback adjustment factor, preset benchmark step size and stage decay coefficient, the theoretical step size is calculated, and the adaptive capability step size is obtained by truncating the theoretical step size to the upper and lower limits.

[0008] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of generating user attractiveness based on user QoS deficit weight, comprising: Based on user QoS deficit weights and preset channel matching, service status, and distance attenuation, a comprehensive attractiveness weight for users to drones is generated. The user attraction of a drone is calculated based on the overall attractiveness weight of the drone to the user.

[0009] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of calculating kinematic damping correction based on user attractiveness and adaptive capability step size, comprising: The system obtains the velocity or displacement state of the drone at the previous moment, performs kinematic damping correction on the drone's velocity or displacement state based on a preset damping correction coefficient, and updates the user attraction and adaptive capability step size based on the correction result to obtain the corrected three-dimensional displacement.

[0010] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of updating the UAV position based on the corrected three-dimensional displacement, comprising: The UAV updates its position based on the corrected 3D displacement. After the UAV position is updated, an airspace boundary check is performed, and the safe distance constraints between UAVs are calculated. If the updated position of the UAV violates the safe distance constraints between UAVs, the UAV is corrected to the legal airspace by projection, truncation, or local rollback.

[0011] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of updating energy consumption and service status based on the updated UAV location, comprising: The propulsion power, single-step propulsion energy consumption, and remaining power of the drone are calculated based on the actual flight speed of the drone. At the same time, the drone trajectory, user speed, load status, joint evaluation value, and energy consumption status at the current moment are recorded to provide feedback for the dynamic deployment of the drone in the next moment.

[0012] Secondly, in order to achieve the above objectives, this invention discloses an air-ground cooperative UAV energy efficiency perception virtual force dynamic deployment system, comprising: The data acquisition module is used to acquire current network status information, which includes: the horizontal position coordinates of the user equipment, the current downlink achievable rate of the user, the location of the ground access point, the three-dimensional location of the UAV air access point, the large-scale channel gain between the user and each access point, the user's current service cluster, the remaining battery power of the UAV, the speed or displacement status of the UAV at the previous moment, and the set of users currently served by the UAV. The data processing module is used to calculate based on the current network status information to obtain the user QoS deficit weight and adaptive capability step size, generate user attraction based on the user QoS deficit weight, and perform kinematic damping correction calculation based on user attraction and adaptive capability step size to obtain the corrected three-dimensional displacement. The update output module is used to update the UAV's position based on the corrected 3D displacement, and to update the energy consumption and service status based on the updated UAV position, thereby enabling the dynamic deployment of the UAV.

[0013] The beneficial effects of this invention are: This invention collects real-time status information from a co-located, non-cellular, massive MIMO network, then constructs a user-centric service cluster, calculates user QoS deficit, UAV payload, and joint state assessment value, and generates the UAV's three-dimensional position update direction based on a multi-source virtual force field. By combining anti-hysteresis disturbance, energy-adaptive step-size control, and kinematic damping correction mechanisms, closed-loop adjustment of the UAV position update is achieved, thereby realizing real-time dynamic deployment that balances user fairness, UAV load balancing, trajectory smoothness, and energy constraints. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1This is a schematic diagram of the method flow of the present invention; Figure 2 This is a schematic diagram of the design process of the method of the present invention; Figure 3 This is a schematic diagram of the overall modules within an embodiment of the present invention; Figure 4 This is a schematic diagram of the system structure of the present invention; Figure 5 This is a schematic diagram showing the cumulative propulsion energy consumption comparison results of the present invention; Figure 6 This is a schematic diagram showing the long-term unit energy efficiency comparison results of this invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example 1: like Figure 1 As shown, a method for dynamically deploying virtual forces for energy efficiency perception of air-ground cooperative UAVs includes the following steps: S101: Obtain current network status information, which includes: horizontal position coordinates of user equipment, current downlink achievable rate of user, location of ground access point, three-dimensional location of UAV air access point, large-scale channel gain between user and each access point, current service cluster of user, remaining battery power of UAV, speed or displacement status of UAV at the previous moment, and set of service users currently served by UAV. Based on the current network state information, calculations are performed to obtain the user QoS deficit weight, UAV effective load state, network joint state evaluation value, and adaptive capability step size. User attraction is generated based on user QoS deficit weight, channel matching state, service state, and distance attenuation. Topological security repulsion force is generated in combination with the safe distance constraint between UAVs. Air-ground cooperative spring force is generated in combination with the location of the ground access point. Load response adjustment force and directional bias diversion force are generated in combination with the effective load state of UAVs. Altitude potential energy adjustment force is generated in combination with the altitude state of UAVs. The forces are synthesized to obtain the corresponding three-dimensional balance virtual force of UAVs. Kinematic damping correction calculation is performed based on the three-dimensional balance virtual force, adaptive capability step size, and the speed or displacement state of UAVs at the previous moment to obtain the corrected three-dimensional displacement. Specifically, the initialization process of information processing includes: Initialize the air-ground cooperative network: Define the target service area, the set of ground access points, the set of UAV aerial access points, and the set of user equipment. Assume the network coverage area is a two-dimensional ground region, and the set of ground access points is represented as follows: The set of drone air access points is represented as: The set of user equipment is represented as: The position of the k-th user is represented as: The location of the g-th ground access point is represented as: The three-dimensional position of the m-th UAV is represented as: Simultaneously initialize the drone's safe distance. Legal flight altitude range Maximum flight speed Maximum acceleration Initial battery capacity Reference step size Stagnation detection window And various virtual force control parameters.

[0017] Collect current network status: Within each time step t, the central control unit periodically collects the current network status, including user location. User business needs, current downlink achievable speed Ground access point location, drone location Large-scale channel gain between users and access points Remaining battery power of the drone Velocity at the previous moment And the current user base of drone services.

[0018] In dynamic scenarios, the positions of the user and the drone change over time. The drone's position update can be approximated as: in, For discrete time steps, This represents the instantaneous speed of the drone at the current moment.

[0019] Build a user-centric service cluster Based on the large-scale link gain between users and ground access points and drone access points, a service access point cluster is dynamically constructed for each user. Let user... The service clusters are: Define binary service association variables: Here, 'l' represents any ground access point or drone access point. The purpose of this step is to select the access point that contributes most to the user based on the real-time link status, thereby reducing the fronthaul pressure and signaling overhead brought about by the fully connected non-cellular architecture.

[0020] S102: Calculate the user QoS deficit weight and adaptive capability step size based on the current network state information, generate user attraction based on the user QoS deficit weight, and perform kinematic damping correction calculation based on user attraction and adaptive capability step size to obtain the corrected three-dimensional displacement. The UAV position is updated based on the corrected 3D displacement, and the updated UAV position is verified by airspace boundary constraints and safe distance constraints between UAVs. Based on the updated UAV position, the user service cluster, user rate, UAV payload, network joint state assessment value, propulsion energy consumption and UAV remaining power are updated. The updated service status, load status, energy consumption status and motion status are used as feedback information for dynamic deployment in the next moment, realizing real-time dynamic deployment of the virtual force of energy efficiency perception of air-ground cooperative UAVs.

[0021] The calculation process for user QoS deficit weight is as follows: The user's QoS deficit weight is calculated based on the current downlink achievable rate within the current network status information and using the inverse rate normalization method.

[0022] The magnitude of a user's QoS deficit weight is inversely proportional to the magnitude of the user's current downlink achievable rate.

[0023] The specific process is as follows: like Figure 2 As shown, the QoS deficit is calculated based on the current user downlink rate. For users... The lower the rate, the worse the communication service quality, and therefore the higher its weight should be in subsequent virtual gravity calculations. Therefore, the QoS deficit weight can be defined using the inverse rate normalization method: in, This is a smoothing constant used to avoid... Too low a weight can cause numerical instability. This weighting increases the pull on drones from low-rate users, edge users, and users with weak coverage, thereby raising the minimum user rate of the system.

[0024] The system further defines the minimum user speed for the entire network as: This metric is used to measure the fairness of communication on the user side.

[0025] This also includes calculating the drone's payload: For each drone The system calculates the payload based on the degree of dependence of its service users on the drone link, rather than simply counting the number of service users. Let's assume a drone... The user set of the service is At the same time, it covers users The set of ground access points is Then the payload of the drone m can be expressed as: in, Indicates user With drones Large-scale channel gain between Indicates user ground access point Large-scale channel gain between.

[0026] When a user is simultaneously covered by strong ground access points, that user contributes less to the drone's load; conversely, when a user primarily relies on drone links, that user contributes more to the drone's load. This more accurately reflects the actual service pressure of drones in air-ground cooperative networks.

[0027] Calculate the network-side load balancing index After obtaining the payload of all drones, the system uses the Jain fairness index to evaluate the load balancing level among the drones: in, .when The closer the value is to 1, the more balanced the load distribution among the drones; when the value is low, it indicates that some drones are at risk of overload, while other drones may be in a light-load or idle state.

[0028] Furthermore, a joint state evaluation function is constructed: The system integrates user-side fairness and network-side load balancing into the joint state evaluation function. First, the original joint score is defined: in, The minimum user rate after normalization. and These represent the weights of user-side fairness and drone load balancing in the joint evaluation, respectively.

[0029] To prevent the system from sacrificing low-speed users for superficial load balancing, a safe rate floor is introduced. And construct a soft lower bound penalty mechanism: The joint state assessment value is used not only to evaluate the current deployment effect, but also for subsequent anti-lag judgment and performance feedback adjustment.

[0030] The process of generating user attractiveness based on user QoS deficit weights includes: Based on user QoS deficit weights and preset channel matching, service status, and distance attenuation, a comprehensive attractiveness weight for users to drones is generated. The user attraction of a drone is calculated based on the overall attractiveness weight of the drone to the user.

[0031] Specifically, this includes: for each drone The system calculates the attractiveness from all users. (Assuming a drone...) With users The direction vector in the horizontal plane is: The horizontal distance is: The overall attractiveness of drones to users is determined by a combination of factors including QoS deficit, channel matching, service status, and distance attenuation. The channel matching factor can be expressed as: The service status exploration factor can be represented as: in, This is intended to moderately reduce the repeated towing of the drone by existing users and guide the drone to expand into areas with insufficient coverage.

[0032] The distance attenuation factor can be expressed as: in, To effectively control the radius, To adjust the parameters.

[0033] Therefore, drones The user appeal is: in, Based on the gravitational gain coefficient, To prevent division by zero of extremely small constants.

[0034] It also includes constructing topological security repulsion forces. To prevent drones from getting too close, the system calculates the safety repulsion force between them. When any two drones... and Horizontal distance between Less than the safe distance threshold At this time, a nonlinear repulsive force is triggered: drones The total safety repulsion force is: This repulsive force is used to meet the collision avoidance requirements of drones and reduce coverage redundancy and co-frequency interference caused by excessive clustering in local airspace.

[0035] Constructing an air-ground coordinated spring force To maintain the cooperative topology between the airborne UAV and the ground access point, the system treats the ground access point as a fixed virtual anchor point and introduces a cooperative air-ground spring force. The spring force acting on the UAV m can be expressed as: in, This is the spring stiffness coefficient. The ideal loiter distance for air-to-ground collaboration. If the drone is too far from the ground network, the spring force will act as an attraction; if the drone gets too close to the ground access point, the spring force will act as a repulsion.

[0036] Perform load-aware horizontal coordination adjustment The system determines the relative congestion status of drones based on their effective load. Let the average load of all drones in the network be: drones The load ratio is: when Greater than the overload threshold At that time, it was explained that the drone The system is in an overload state. The system's load response factor is: in, This is the load adjustment strength coefficient. The gravitational force after load adjustment is: This mechanism can prevent overloaded drones from continuing to gather in hotspot areas.

[0037] Furthermore, for overloaded drones, the system introduces a directional bias factor based on the alternative coverage capabilities of the ground access point for its service users: When a user is well covered by a ground access point, the proportion of their attraction to overloaded drones decreases, thereby guiding overloaded drones to smoothly leave the hotspot center and achieving horizontal diversion.

[0038] Perform high potential energy regulation In addition to horizontal adjustment, the system also utilizes the drone's altitude freedom to alleviate localized service congestion. When the drone is overloaded, the system generates an upward adjustment force in the vertical direction to raise it appropriately and expand the coverage area; when the load returns to normal, a spring return force brings it back to near its equilibrium altitude.

[0039] drones The high potential energy can be expressed as: in, To adjust the gain for overload, For highly regressive coefficients, To balance flight altitude, this step expands the UAV's side load balancing from two-dimensional horizontal adjustment to three-dimensional spatial adjustment.

[0040] Synthetic three-dimensional equilibrium virtual force Taking into account user appeal, topological security repulsion, air-to-ground cooperative spring force, load diversion force, and altitude potential capability, the UAV is obtained Total virtual force in three-dimensional space: The resultant horizontal force is: This combined force serves as the direct driving force for drone location updates.

[0041] Anti-hysteresis judgment is performed based on the joint state evaluation function. The system calculates the joint state evaluation value after each round of position updates. If continuous Within a given time step, the joint state assessment value improves below a threshold. If the system enters a state of partial stagnation, then it is determined that the system has entered a state of partial stagnation When the stagnant state continues to meet the preset window length, the system triggers a differentiated anti-stagnant disturbance mechanism. For overloaded drones, the system applies a large displacement in a direction away from the centroid of its serving users; for lightly loaded drones, the system guides them to the vicinity of the user with the lowest current rate; for load-balanced drones, only a small random disturbance is applied to preserve the existing reasonable topology.

[0042] Setting up overloaded drones The service user core is: The anti-hysteresis disturbance of an overloaded drone can then be expressed as: in, This is the step size for dynamically decaying the perturbation.

[0043] The calculation process for the adaptive capability step size is as follows: The energy scaling factor is calculated based on the drone's remaining battery power and a preset safe battery power threshold, according to the current network status information. A feedback adjustment factor is generated based on the user's current downlink achievable rate according to the current network status information. Based on the energy scaling factor, feedback adjustment factor, preset benchmark step size and stage decay coefficient, the theoretical step size is calculated, and the adaptive capability step size is obtained by truncating the theoretical step size to the upper and lower limits.

[0044] The specific calculation process includes: In dynamic scenarios, the system dynamically adjusts the drone's maneuverability based on its remaining battery power. (Defining the drone...) The normalized remaining power ratio is: Let the safe power threshold be Then the energy scaling factor is: when At that time, the drone has full maneuverability; when At that time, its step size gradually decreases as the remaining power decreases.

[0045] Furthermore, the system can introduce a performance feedback factor based on the minimum user rate variation. Let: The feedback adjustment factor can then be expressed as: in, When indicating performance improvement, the step size should be increased appropriately. This indicates the compression step size when performance degrades.

[0046] Simultaneously, a stage attenuation coefficient is introduced: The final theoretical step size is: The actual execution step size is obtained by truncation through upper and lower limits: in, This is used to satisfy the maximum flight speed constraint.

[0047] The process of calculating kinematic damping correction based on user attractiveness and adaptive capability step size includes: The system obtains the velocity or displacement state of the drone at the previous moment, performs kinematic damping correction on the velocity or displacement state of the drone at the previous moment based on the preset damping correction coefficient, and updates the user attraction and adaptive capability step size based on the correction result to obtain the corrected three-dimensional displacement.

[0048] Specifically, this includes: to reduce drone trajectory oscillations and frequent back-and-forth movements, the system incorporates the velocity or displacement state from the previous moment to perform damping correction on the current virtual force direction.

[0049] First, the direction of the horizontal virtual force is normalized: In an ideal state with no inertia, the desired horizontal displacement is: After introducing kinematic damping, the actual horizontal displacement is: in, This is the damping coefficient. This mechanism prevents the drone from drastically changing direction due to instantaneous virtual force, instead maintaining trajectory continuity while responding to network changes.

[0050] The vertical displacement can be truncated based on the potential energy at height: in, As a reference force scalar, This represents the maximum step size for height variation.

[0051] S103: Update the UAV position based on the corrected three-dimensional displacement, and update the energy consumption and service status based on the updated UAV position to achieve dynamic deployment of the UAV.

[0052] The process of updating the UAV's position based on the corrected 3D displacement includes: The UAV updates its position based on the corrected 3D displacement. After the UAV position is updated, an airspace boundary check is performed, and the safe distance constraints between UAVs are calculated. If the updated position of the UAV violates the safe distance constraints between UAVs, the UAV is corrected to the legal airspace by projection, truncation, or local rollback.

[0053] The specific calculation process includes: The drone updates its position based on the corrected 3D displacement: After the update, the system performs airspace boundary checks: Simultaneously check the safety distance constraints between drones: If the position update result violates the boundary or safe distance constraints, it is corrected to the legal airspace by projection, truncation or local rollback.

[0054] The process of updating energy consumption and service status based on the updated drone location includes: The propulsion power, single-step propulsion energy consumption, and remaining battery power of the drone are calculated based on the actual flight speed of the drone. At the same time, the drone trajectory, user speed, load status, joint evaluation value, and energy consumption status at the current moment are recorded to provide feedback for the dynamic deployment of the drone in the next moment.

[0055] Specifically, this includes: the system calculating propulsion power based on the actual flight speed of the UAV. Let the UAV... The velocity at the current time step is: The propulsion power of a rotary-wing UAV can be expressed as: in, For type resistance power constant, To induce a power constant, The rotor blade tip velocity, For hovering induced speed, The drag coefficient of the fuselage. air density, For rotor rigidity, The area of ​​the rotor's rotating disk.

[0056] The energy consumption for a single-step propulsion is: The drone's remaining battery power has been updated to: Simultaneously, it records the drone's trajectory, user speed, load status, joint evaluation value, and energy consumption status at the current moment, providing feedback for dynamic deployment in the next time step.

[0057] In dynamic scenarios, the joint state assessment of the system further considers the energy consumption of the UAV, which is reflected in the dynamic fitness function. : In actual dynamic deployment, The value serves as the feedback basis for the adjustment of the three-dimensional balance virtual force and the update of the UAV position, enabling the UAV to balance service continuity, load balancing and energy consumption optimization when responding to changes in user distribution and business needs, thereby achieving long-term stable operation and high energy efficiency of the air-ground collaborative UAV network.

[0058] In the next time step, the system re-collects user location, link status, service cluster, drone load, and remaining battery power, and repeats the above steps. Through this closed-loop process, the drone swarm can continuously adjust its three-dimensional position in scenarios of user movement, service changes, and link fluctuations, achieving real-time dynamic deployment that prioritizes user fairness, network load balancing, flight safety, trajectory smoothing, and long-term energy efficiency.

[0059] Specifically, the present invention will be further illustrated below through embodiments: like Figure 3 As shown, the present invention comprises the following core functional modules.

[0060] 1. Network Status Acquisition Module The network status acquisition module is used to periodically acquire the current status information of the air-ground cooperative network, including the horizontal position coordinates of user equipment, the user's current downlink reachable rate, the location of the ground access point, the three-dimensional location of the UAV's air access point, the large-scale channel gain between the user and each access point, the user's current service cluster, the remaining battery power of the UAV, the speed or displacement status of the UAV at the previous moment, and the set of users currently served by the UAV. This module can be deployed in the central control unit, or it can be implemented by the central control unit in collaboration with edge computing nodes.

[0061] 2. User-Centered Service Cluster Construction Module This module dynamically constructs a cluster of service access points for each user based on the large-scale link gain between the user and each access point. For any given user, the system selects several ground access points and UAV access points as cooperative service nodes for that user based on indicators such as path loss, shadow fading, link gain, or received signal strength. Service relationships can be represented by a binary association variable. This variable is set to 1 when an access point participates in the user's service; otherwise, it is set to 0. The purpose of this module is to reduce the signaling overhead and fronthaul pressure brought by the fully connected non-cellular architecture, while providing a service relationship foundation for subsequent payload calculation and virtual force construction.

[0062] 3. User-side fairness assessment module This module is used to calculate user-side communication fairness metrics. The system uses the lowest user rate among all network users as the primary measure of user-side fairness. The higher the lowest user rate, the stronger the communication guarantee capability for disadvantaged users. In this invention, user-side fairness is not only used for final performance evaluation but also directly participates in virtual force construction. For users with currently low rates, the system considers them as users with large QoS deficits and assigns them higher weight in the drone attractiveness calculation.

[0063] 4. UAV Payload Calculation Module This module calculates the effective load of each UAV. Unlike traditional methods that calculate load based on the number of users served, this invention calculates UAV load based on the relative dependence of users on the UAV link and the ground access point link. For a given UAV, its effective load is obtained by summing the dependence of its serving users on the UAV's channel gain. If a user is also covered by strong ground access points, that user's contribution to the UAV load is lower; if a user mainly relies on UAVs for service, their load contribution is higher. This module can more realistically reflect the actual service pressure of UAVs in air-to-ground cooperative networks.

[0064] 5. Joint State Assessment Module This module is used to comprehensively evaluate the current deployment status. The system weights and fuses the normalized minimum user rate and the drone payload Jain fairness index to obtain a joint state evaluation value. To avoid sacrificing basic communication guarantee capabilities in pursuit of load balancing, this invention further introduces a soft lower bound penalty mechanism. When the minimum user rate is lower than a preset safe rate lower bound, a secondary penalty is applied to the joint state evaluation value, ensuring that the system prioritizes the basic communication needs of low-rate users.

[0065] 6. Balanced Virtual Force Generation Module This module is the core module of the invention, used to generate the three-dimensional resultant force required for UAV position updates. It includes an adaptive gravity generation unit based on QoS deficit, a channel gain matching weighting unit, a service state exploration weighting unit, a distance attenuation weighting unit, a UAV inter-UAV safety repulsion force generation unit, an air-to-ground cooperative spring force generation unit, a load response amplitude scaling unit, a direction offset diversion unit, and an altitude potential energy adjustment unit.

[0066] 7. Anti-disturbance module This module addresses potential local stagnation issues during virtual force evolution. When the joint state evaluation value improves below a preset threshold after several consecutive iterations, the system is deemed to have entered a stagnant state, triggering a differentiated perturbation strategy. For overloaded drones, the system significantly displaces them away from the centroid of serving users to disrupt local equilibrium within hotspot areas. For lightly loaded drones, the system directionally guides them to the vicinity of the user with the lowest current rate to improve idle resource utilization. For load-balanced drones, only minor random adjustments are applied to preserve existing effective topology.

[0067] 8. Energy Adaptive Step Size Control Module This module adjusts the maneuver permissions of drones based on their remaining battery power in dynamic scenarios. When a drone has sufficient battery power, the system allows it to fully respond to virtual forces and perform large-scale deployment adjustments; when the drone's battery power falls below a safe threshold, the system gradually reduces its movement step size, causing it to transition from a high-frequency maneuver state to a local fine-tuning or relatively stationary state. This module can prevent individual drones from prematurely running out of power due to frequent maneuvers, thereby extending the overall service time of the cluster.

[0068] 9. Kinematic Damping and Trajectory Correction Module This module reduces drone trajectory oscillations in dynamic user scenarios. When calculating the drone's current position update, the system incorporates its previous velocity or displacement state to apply damping correction to the current virtual force direction. Through this module, the drone avoids drastic turns by completely following the instantaneous force direction, instead maintaining a continuous and smooth trajectory while responding to network changes.

[0069] To further verify the technical advantages of the method described in this embodiment compared to existing UAV deployment methods, the method described in this embodiment (Dynamic Energy Efficient Virtual Force, DE2VF) was compared with existing non-cellular discrete genetic algorithm DGA-CF, non-cellular distributed particle swarm optimization algorithm DPSO-CF, and non-cellular hybrid sparrow search algorithm NSSA-CF in the same air-ground cooperative non-cellular large-scale MIMO dynamic simulation environment. All comparison methods were run under the same three-dimensional spatial constraints, the same user mobility model, the same channel update mechanism, and the same service cluster reconstruction rules, and the minimum user rate, UAV load balancing index, joint state assessment value, cumulative propulsion energy consumption, and long-term unit energy efficiency were uniformly used as evaluation indicators.

[0070] In a specific verification scenario, the ground user's location changes continuously over time. The system periodically refreshes the user's location, link status, and dynamic service clusters within multiple consecutive time steps, and calculates the user's downlink rate, drone payload, and flight propulsion energy consumption based on the drone deployment location output by each algorithm. The maximum drone flight speed is set to 10 m / s, the maximum effective battery capacity is set to 1000 kJ, the damping coefficient is set to 0.15, and the reference step size is set to 26 m. Figure 5 The dynamic time-series tracking stability comparison results show that DGA-CF, DPSO-CF, and NSSA-CF all exhibit significant fluctuations in minimum user rate, load balancing index, and joint state evaluation value when user locations continuously change. In particular, the minimum user rate tends to decrease and the joint state evaluation value tends to drop in the early stages of dynamic time step switching. This indicates that the traditional deployment method based on candidate solution search is closer to independently re-optimizing the network snapshot at each moment, making it difficult to fully inherit the topology state of the previous moment, which can easily lead to significant reconstruction of UAV locations and fluctuations in service status.

[0071] Further integration Figure 5 The comparison of cumulative propulsion energy consumption shows that DGA-CF, DPSO-CF, and NSSA-CF, lacking unified constraints on trajectory smoothness and remaining energy status across continuous time dimensions, are prone to repetitive maneuvers and ineffective backtracking during multi-time-step deployment, leading to a rapid increase in cumulative propulsion energy consumption. The method described in this embodiment limits and modulates the actual displacement amplitude of the UAV at each time step through remaining power maneuver permission allocation, performance feedback adjustment, and phase decay control. This ensures the UAV has the necessary responsiveness when the network topology is unstable, and automatically transitions to a low-energy fine-tuning state when the topology gradually stabilizes or remaining energy decreases, thereby reducing the additional energy consumption caused by frequent maneuvers.

[0072] like Figure 6As shown, the long-term unit energy efficiency comparison results further demonstrate that the unit energy efficiency of existing comparison methods decreases rapidly with continuous operation time, indicating that the short-term performance gains obtained through frequent location reconfiguration cannot offset the long-term costs caused by the cumulative increase in propulsion energy consumption. In contrast, the unit energy efficiency of the method described in this embodiment remains at a high level throughout, and the downward trend is more gradual, indicating that it can maintain good user service quality and UAV load balancing with lower energy consumption during long-term continuous operation. Therefore, the method described in this embodiment not only improves service continuity and trajectory smoothness during dynamic deployment but also reduces ineffective maneuver energy consumption, improving the long-term energy efficiency performance of air-ground cooperative UAV networks.

[0073] Example 2: To achieve the above objective, such as Figure 4 As shown, based on Embodiment 1, this invention discloses an air-ground cooperative UAV energy efficiency perception virtual force dynamic deployment system, comprising: The data acquisition module 11 is used to acquire current network status information, which includes: the horizontal position coordinates of the user equipment, the current downlink reachable rate of the user, the location of the ground access point, the three-dimensional location of the UAV air access point, the large-scale channel gain between the user and each access point, the user's current service cluster, the remaining battery power of the UAV, the speed or displacement status of the UAV at the previous moment, and the set of service users currently undertaken by the UAV. Data processing module 12 is used to calculate based on the current network status information to obtain the user QoS deficit weight and adaptive capability step size, generate user attraction based on the user QoS deficit weight, and perform kinematic damping correction calculation based on user attraction and adaptive capability step size to obtain the corrected three-dimensional displacement. The update output module 13 is used to update the UAV position based on the corrected three-dimensional displacement, and update the energy consumption and service status based on the updated UAV position to realize the dynamic deployment of UAV.

[0074] Based on the same inventive concept, this invention also provides a computer device, comprising: one or more processors, and a memory for storing one or more computer programs; the programs include program instructions, and the processor executes the program instructions stored in the memory. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, used to implement one or more instructions, specifically for loading and executing one or more instructions stored in a computer storage medium to implement the above-described method.

[0075] It should be further explained that, based on the same inventive concept, the present invention also provides a computer storage medium storing a computer program, which, when executed by a processor, performs the above-described method. This storage medium can be any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In the present invention, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0076] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0077] The foregoing has shown and described the basic principles, main features, and advantages of this disclosure. Those skilled in the art should understand that this disclosure is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of this disclosure. Various changes and modifications can be made to this disclosure without departing from its spirit and scope, and all such changes and modifications fall within the scope of this disclosure as claimed.

Claims

1. A method for dynamically deploying virtual forces for energy efficiency perception of air-ground cooperative UAVs, characterized in that, The method includes the following steps: Obtain current network status information, which includes: the horizontal position coordinates of the user equipment, the current downlink achievable rate of the user, the location of the ground access point, the three-dimensional location of the UAV air access point, the large-scale channel gain between the user and each access point, the user's current service cluster, the remaining battery power of the UAV, the speed or displacement status of the UAV at the previous moment, and the set of users currently served by the UAV. The user QoS deficit weight and adaptive capability step size are calculated based on the current network state information. The user attraction is generated based on the user QoS deficit weight. The kinematic damping correction is calculated based on the user attraction and adaptive capability step size to obtain the corrected three-dimensional displacement. The drone's position is updated based on the corrected 3D displacement, and the energy consumption and service status are updated based on the updated drone position, enabling dynamic deployment of the drone.

2. The method for dynamic deployment of virtual force for energy efficiency perception of air-ground cooperative UAVs according to claim 1, characterized in that, The calculation process for the user QoS deficit weight is as follows: The user's QoS deficit weight is calculated based on the current downlink achievable rate within the current network status information and using the inverse rate normalization method.

3. The method for dynamic deployment of virtual force for energy efficiency perception of air-ground cooperative UAVs according to claim 2, characterized in that, The magnitude of the user's QoS deficit weight is inversely proportional to the magnitude of the user's current downlink achievable rate.

4. The method for dynamic deployment of virtual force for energy efficiency perception of air-ground cooperative UAVs according to claim 1, characterized in that, The calculation process for the adaptive capability step size is as follows: The energy scaling factor is calculated based on the drone's remaining battery power and a preset safe battery power threshold, according to the current network status information. A feedback adjustment factor is generated based on the user's current downlink achievable rate according to the current network status information. Based on the energy scaling factor, feedback adjustment factor, preset benchmark step size and stage decay coefficient, the theoretical step size is calculated, and the adaptive capability step size is obtained by truncating the theoretical step size to the upper and lower limits.

5. The method for dynamic deployment of virtual force for energy efficiency perception of air-ground cooperative UAVs according to claim 1, characterized in that, The process of generating user attractiveness based on user QoS deficit weight includes: Based on user QoS deficit weights and preset channel matching, service status, and distance attenuation, a comprehensive attractiveness weight for users to drones is generated. The user attraction of a drone is calculated based on the overall attractiveness weight of the drone to the user.

6. The method for dynamically deploying virtual forces for energy efficiency perception of air-ground cooperative UAVs according to claim 1, characterized in that, The process of calculating kinematic damping correction based on user attractiveness and adaptive capability step size includes: The system obtains the velocity or displacement state of the drone at the previous moment, performs kinematic damping correction on the drone's velocity or displacement state based on a preset damping correction coefficient, and updates the user attraction and adaptive capability step size based on the correction result to obtain the corrected three-dimensional displacement.

7. The method for dynamic deployment of virtual force for energy efficiency perception of air-ground cooperative UAVs according to claim 1, characterized in that, The process of updating the UAV position based on the corrected three-dimensional displacement includes: The UAV updates its position based on the corrected 3D displacement. After the UAV position is updated, an airspace boundary check is performed, and the safe distance constraints between UAVs are calculated. If the updated position of the UAV violates the safe distance constraints between UAVs, the UAV is corrected to the legal airspace by projection, truncation, or local rollback.

8. The method for dynamically deploying virtual forces for energy efficiency perception of air-ground cooperative UAVs according to claim 1, characterized in that, The process of updating energy consumption and service status based on the updated drone location includes: The propulsion power, single-step propulsion energy consumption, and remaining battery power of the drone are calculated based on the actual flight speed of the drone. At the same time, the drone trajectory, user speed, load status, joint evaluation value, and energy consumption status at the current moment are recorded to provide feedback for the dynamic deployment of the drone in the next moment.

9. A dynamic deployment system for energy-efficiency sensing virtual force of an air-ground cooperative UAV, employing the dynamic deployment method for energy-efficiency sensing virtual force of an air-ground cooperative UAV as described in any one of claims 1 to 8, characterized in that, include: The data acquisition module is used to acquire current network status information, which includes: the horizontal position coordinates of the user equipment, the current downlink achievable rate of the user, the location of the ground access point, the three-dimensional location of the UAV air access point, the large-scale channel gain between the user and each access point, the user's current service cluster, the remaining battery power of the UAV, the speed or displacement status of the UAV at the previous moment, and the set of users currently served by the UAV. The data processing module is used to calculate based on the current network status information to obtain the user QoS deficit weight and adaptive capability step size, generate user attraction based on the user QoS deficit weight, and perform kinematic damping correction calculation based on user attraction and adaptive capability step size to obtain the corrected three-dimensional displacement. The update output module is used to update the UAV's position based on the corrected 3D displacement, and to update the energy consumption and service status based on the updated UAV position, thereby enabling the dynamic deployment of the UAV.

10. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, The memory stores a computer program that can run on the processor. When the processor loads and executes the computer program, it employs the method for dynamic deployment of virtual force for energy efficiency perception of air-ground cooperative UAVs as described in any one of claims 1 to 8.

Images