A lightweight slice resource configuration method and system for unmanned aerial vehicle assisted network

Abstract

The present application belongs to the technical field of unmanned aerial vehicle, disclose a kind of light slice resource configuration method for unmanned aerial vehicle assisted network, the present application is by joint optimization channel allocation, power allocation and unmanned aerial vehicle deployment, to maximize minimum user rate.UAWN slice problem is modeled as non-convex mixed integer nonlinear programming.Considering the limited computing power of unmanned aerial vehicle, a low-complexity decomposition framework is proposed, combining problem decomposition, integer relaxation and batch coordinate descent method to solve the problem.In addition, an optimal rounding algorithm is designed to obtain the relaxation solution, ensuring the feasibility and optimality of the rounded solution.To verify the effectiveness of the scheme, a large number of numerical experiments are carried out.The simulation results show that the convergence speed and performance of the scheme are superior to some existing algorithms.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) technology, and particularly relates to a lightweight slicing resource configuration method and system for UAV-assisted networks. Background Technology

[0002] Non-terrestrial networks (NTNs) are considered one of the pillars of the upcoming 6G network, capable of serving a variety of critical applications such as environmental monitoring, public safety, disaster search and rescue, and procurement delivery. In particular, unmanned aerial vehicle (UAWN) networks have attracted considerable attention from academia and industry due to their flexible deployment, rapid relay, and line-of-sight (LoS) links. In recent years, with the rise of the low-altitude economy, the diversity of UAWN services has shown a rapid growth trend. Since a single network architecture is not suitable for all situations, it is unrealistic to expect a single UAWN to meet these diverse service demands. Network slicing is the creation of virtualized independent logical networks on a shared physical infrastructure to meet the specific needs of different services. By introducing network slicing technology into UAWNs, it is possible to avoid deploying separate UAWNs for each service type, thereby rationally allocating UAWN resources.

[0003] Because network slicing eliminates the need to deploy separate networks for each service type, it can reduce the number of drones while providing differentiated services. This reduction in drone numbers proportionally lowers the capital and operating costs of the UAWN, and also means reduced interference and collisions between drones, simplifying drone coordination within the network. Furthermore, technologies supporting network slicing, such as Network Functions Virtualization (NFV) and Software Defined Networking (SDN), can significantly reduce the amount of physical network equipment installed on drones, allowing them to dedicate more energy to data transmission and thus extend their service life.

[0004] Fully leveraging the advantages of network slicing requires addressing several challenges. First, airborne base stations (BSs) must consider both UAV deployment and network resource allocation to adapt to the ever-changing three-dimensional network. Second, given the limited computing power and battery capacity of UAVs, low-complexity network slicing schemes must be designed. Finally, guaranteeing the Service Level Agreement (SLA) for slices in resource-constrained UAWNs is also a challenging task. However, besides the scarcity of research addressing these issues, existing research has failed to provide a comprehensive solution to overcome all these challenges. While some recent studies have jointly considered UAV deployment and resource allocation, these solutions are complex, slow to converge, and lack SLA guarantees, limiting their application in UAV networks.

[0005] First, existing solutions are excessively complex, exceeding the computational capabilities of UAVs. Second, while some researchers have employed decomposition techniques to simplify the problem, these decompositions are often heuristic and cannot guarantee the equivalence of the decomposed problem with the original problem, resulting in poor performance. Third, these studies lack SLA modeling, leading to unacceptable SLA violations for slice users. In contrast, the proposed decomposition method is theoretically sound, and in a sense, the decomposed problem is equivalent to the original problem. Furthermore, the model incorporates SLA constraints for network slicing, ensuring that the proposed solution guarantees the Quality of Service (QoS) for users. Finally, the proposed solution framework is lightweight, converging within milliseconds, making it suitable for UAVs with limited computational resources in rapidly changing environments.

[0006] Based on the above analysis, the problems and shortcomings of the existing technology are as follows:

[0007] (1) The existing solutions are too complex and exceed the computing power of the UAV.

[0008] (2) Although some researchers have used certain decomposition techniques to simplify the problem, these decompositions are usually heuristic and cannot guarantee the equivalence of the decomposed problem with the original problem, resulting in poor performance.

[0009] (3) These studies lack SLA modeling, which leads to SLA violations that are unacceptable to slice users. Summary of the Invention

[0010] To address the problems existing in the prior art, this invention provides a lightweight slicing resource configuration method for UAV-assisted networks.

[0011] This invention is implemented as follows: A lightweight slicing resource configuration method for UAV-assisted networks includes:

[0012] Step 1: Divide the problem into two equivalent parts: a resource allocation problem and a drone deployment problem;

[0013] Step 2: By analyzing the properties of the subproblems, derive the explicit solutions to the subproblems, and then derive the explicit expression of the main problem.

[0014] Step 3: Prove that the relaxation principal problem is a pseudo-concave problem, and its stationary point corresponds to the global optimal solution. Based on this, use the KKT conditions to find the global optimal solution using a lightweight decomposition algorithm, and design the optimal rounding algorithm to minimize the rounding loss of the relaxation solution.

[0015] Furthermore, the derivation method is as follows:

[0016] Basis network model

[0017] First, let's focus on downlink transmission, where the drone is equipped with a maximum transmission power of P. max The antenna, the bandwidth available to the drone is denoted as B. tot Assuming the drone's projected position on the horizontal plane is the center of the drone network; and letting θ be the beamwidth and h be the drone's height; then the following coverage conditions must be met:

[0018]

[0019] In the above formula, R is the radius of the UAWN; considering that the flight capability of the UAWN is limited, its altitude is restricted by the following conditions:

[0020]

[0021] In the above formula, c represents the flight speed of the drone. T represents the drone's flight altitude in the previous time slot. s The duration of each time slot in the network;

[0022] Network slicing model

[0023] use Let i represent the set of network slices, and let i be the set of user devices. u ij Let u represent the j-th user device in slice i. ij The two-dimensional coordinates are as follows User equipment u ij The square distance between the drone and the drone is in Let g0 be the channel power gain at a reference distance of one meter; according to existing literature, user equipment u ijThe channel power gain between the drone and the other device can be modeled as follows:

[0024]

[0025] Where α is the path loss exponent; in This represents the small-scale fading channel coefficients of the Rayleigh distribution, that is... Follows a non-central chi-square distribution;

[0026] Use B i Let x represent the channel bandwidth of slice i; it should be noted that due to service differentiation, the channel bandwidth of different slices is different; use a non-negative integer x. ij Indicates allocation to user equipment u ij The number of channels; obviously, x ij The following capacity constraints must be met:

[0027]

[0028] Among them, drones deliver to user equipment u ij The transmitted power is denoted as p. ij The following capacity constraints must be met:

[0029]

[0030] To guarantee the SLA, each slice should provide a guaranteed data rate to subscribers, i.e.:

[0031]

[0032] Where, σ 2 It is the power of the background noise. This is the minimum rate guaranteed by slice i; in addition, each user equipment can request a customized service rate. Considering the tractability of the problem, the constraint is modeled as follows:

[0033]

[0034] in, It is a positive integer and satisfies

[0035] UAWN network slicing model

[0036] The UAWN network slicing problem can be described as follows:

[0037]

[0038] st(1)-(2),(4)-(7)(9)

[0039] h min ≤h≤h max (10)

[0040]

[0041] Among them, h min h is the minimum permissible flight altitude for drones. max Given the maximum permissible flight altitude for drones, the problem can be equivalently divided as follows:

[0042] USP decomposition

[0043] The objective function of USP is denoted as f(x,p,h) = min i,j f ij (x ij ,p ij The so-called problem partitioning refers to projecting both the objective function and constraint set of the USP onto the (x,p) axis. After projecting the USP onto the (x,p) axis, the following main problem can be obtained:

[0044]

[0045] st(x,p)∈V(14)

[0046] Where X and P are defined as follows: X:={x|x satisfies (4) and (7)}, and P:={p|p satisfies (5) and (11)}; set V is defined as:

[0047] V: = {p | there exists h∈H} B Make p satisfy (6)} (15)

[0048] In the above formula, H B It is the interval set defined by constraints (1), (2) and (10), that is in and They are respectively and

[0049] In MP, v(x,p) is defined as a subproblem:

[0050]

[0051] Explicit solutions to subproblems

[0052] The explicit solution to SP can be directly derived. This solution is independent of x and p, and the following theorem about SP can be derived:

[0053] For each fixed pair If the corresponding SP is feasible, then its optimal solution must be

[0054] Explicit expression of the main problem and its relaxation

[0055] Let n represent the index pair (i,j). This refers to all user equipment in the network, i.e. The number of user devices in the system is represented by N. The relaxed MP can be represented as the following problem:

[0056]

[0057] std T x≤B tot (19)

[0058] 1 T p≤P max (20)

[0059] x≥ x ,p≥ p (twenty one)

[0060] Where d and 1 are the coefficients of constraints (4) and (5), respectively. β n For n, the index pair (i,j) contains B i , To make it clearer, let ω(x,p) denote the objective function of MP-Relax; and it can be proven that the inequality constraints of constraints (19) and (20) can be replaced by equality constraints.

[0061] Furthermore, the lightweight decomposition algorithm for finding the global optimum using KKT conditions is as follows:

[0062] Solving the relaxation master problem

[0063] The BCD method will be used to solve MP-Relax. Based on pseudo-convex optimization theory and the characterization of MP-Relax solutions, two subproblems, namely the power allocation subproblem and the sub-channel allocation subproblem, will be solved using the KKT conditions. Then, a BCD algorithm for solving the global optimal solution of MP-Relax will be proposed.

[0064] Power allocator problem

[0065] By fixing a stationary point x on the domain, we can obtain the following power allocation subproblem:

[0066]

[0067] st1 T p≤P max (twenty three)

[0068] P≥ P (twenty four)

[0069] Among them, a n =β n x n It is evident that the power allocator problem is a non-smooth convex programming problem. Since its objective function is strictly convex, it has a unique solution; assuming this optimal solution is (p... * ,z * ),definition:

[0070]

[0071] Sub-channel allocation sub-problem;

[0072] The BCD algorithm for relaxing the master problem.

[0073] Furthermore, the sub-channel allocation sub-problem:

[0074] At the x-coordinate, the following sub-channel allocation sub-problem needs to be solved for a fixed p:

[0075]

[0076] std T x≤B max (28)

[0077] x≥ x (29)

[0078] Where c n =β n ln(1+b n p n );because Sub x The objective function is strictly convex, which indicates the uniqueness of its optimal solution; similarly, assuming the optimal value is t * Assume the optimal solution is It can be deduced that:

[0079]

[0080] Furthermore, the BCD algorithm for the relaxation master problem:

[0081] Based on KKT x and KKT p The algorithm proposes a BCD algorithm based on iterative rules to solve MP-Relax.

[0082] Another object of the present invention is to provide a lightweight slicing resource allocation system for unmanned aerial vehicle (UAV) assisted networks, comprising:

[0083] The partitioning module is used to divide the problem into two equivalent problems: resource allocation and drone deployment.

[0084] The derivation module is used to derive explicit solutions to subproblems and explicit expressions of the main problem by analyzing the properties of subproblems.

[0085] The solution module is used to prove that the relaxation master problem is a pseudo-concave problem, and its stationary point corresponds to the global optimum. Based on this, a lightweight decomposition algorithm is used to find the global optimum using the KKT conditions, and an optimal rounding algorithm is designed to minimize the rounding loss of the relaxation solution.

[0086] Another object of the present invention is to provide a computer device including a memory and a processor, the memory storing a computer program, which, when executed by the processor, causes the processor to perform the steps of the lightweight slice resource allocation method for unmanned aerial vehicle-assisted networks.

[0087] Another object of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of the lightweight slice resource allocation method for unmanned aerial vehicle-assisted networks.

[0088] Another objective of this invention is to provide an information data processing terminal for implementing the lightweight slice resource configuration system for UAV-assisted networks.

[0089] Based on the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solution to be protected by this invention are as follows:

[0090] First, research on existing technical solutions revealed the following technical problems: The complexity of existing solutions is excessive, exceeding the computational capabilities of drones; while some researchers have employed decomposition techniques to simplify the problem, these decompositions are typically heuristic and cannot guarantee the equivalence of the decomposed problem to the original problem, resulting in poor performance; these studies lack SLA modeling, leading to unacceptable SLA violations by slice users. The key to solving these problems lies in comprehensively considering drone deployment, slice resource allocation, and SLA modeling to reduce solution complexity and improve practical applicability.

[0091] This invention studies the UAWN Network Slicing Problem (USP) by considering UAV deployment and resource allocation, aiming to maximize the minimum user rate. Multiple parameters, including UAV flight altitude, power allocation, and channel allocation, are jointly optimized. Research shows that this problem is a large-scale non-convex mixed-integer nonlinear programming problem, extremely difficult to solve. Therefore, based on the problem's structure, a novel batch coordinate descent-based decomposition (BCD) is designed. 2 The framework is used to solve it.

[0092] Unlike existing heuristic-based decomposition methods, this invention is based on rigorous theoretical foundations. First, the USP is decomposed into a Master Problem (MP) and a Subproblem (SP) using projection techniques, where the SP is explicitly solved. It is proven that the relaxed MP is a pseudo-convex problem and solved using the Batch Coordinate Descent-based (BCD) method. By utilizing the Karush-Kuhn-Tucker (KKT) conditions, a corresponding fast algorithm is designed for each subproblem under each coordinate system, with its complexity primarily determined by array sorting. Finally, a rounding algorithm is designed to round the relaxed solution to its integer counterpart, ensuring minimal rounding loss. The main innovations of this invention are as follows:

[0093] (1) The UAWN slicing problem is modeled as a non-convex mixed-integer nonlinear programming problem, and a low-complexity decomposition framework is proposed. This decomposition framework proves the feasibility of designing a log-linear algorithm that can solve large-scale non-convex and non-smooth problems in NTNs within milliseconds.

[0094] (2) The solution found that the throughput maximization problem has good pseudo-concave properties, which can be used to alleviate the non-convexity challenge that is common in NTNs.

[0095] (3) Through a large number of numerical experiments, the scheme reveals the complex relationship between slice resource allocation and UAV deployment, providing guidance for UAV deployment and network slice configuration in UAWN.

[0096] This invention addresses the slicing problem in UAV-assisted networks by proposing a low-complexity optimization framework to jointly optimize resource allocation and UAV deployment. The problem is equivalently divided into a resource allocation problem and a UAV deployment problem. By analyzing the properties of the subproblems, explicit solutions to the subproblems are derived, leading to an explicit expression of the master problem. It is then proven that the relaxed master problem is a pseudo-concave problem, with its stationary point corresponding to the global optimum. Based on this, a lightweight decomposition algorithm using KKT conditions to find the global optimum is proposed. Furthermore, an optimal rounding algorithm is designed to minimize the rounding loss of the relaxed solution. Numerical results demonstrate that, compared to existing algorithms, the proposed algorithm has significant advantages such as faster convergence speed and higher user throughput.

[0097] Secondly, as supporting evidence of the inventiveness of this invention, it is also reflected in the following important aspects:

[0098] (1) The expected benefits and commercial value of the technical solution of this invention after transformation are as follows:

[0099] With the commercialization of network slicing technology and the gradual maturation of research on drone-assisted networks, this invention can be applied to future commercial drone-assisted networks, providing operators with low-complexity customized business solutions and possessing certain commercial value.

[0100] (2) The technical solution of this invention fills a technical gap in the industry both domestically and internationally:

[0101] There is an urgent need for UAV-assisted network slicing technology both domestically and internationally. This invention proposes for the first time a low-complexity decomposition framework, which enables the optimization model of UAV-assisted network slicing to converge quickly and outperforms some known algorithms, providing guidance for UAV deployment and slice resource allocation.

[0102] (3) The technical solution of the present invention solves a technical problem that people have long wanted to solve but have never been able to solve successfully:

[0103] While existing domestic and international solutions for UAV-assisted network slicing can be combined to address UAV deployment and slicing resource allocation, these solutions suffer from high complexity and slow convergence, posing a challenge to the limited computing power of UAVs. This invention reduces the complexity of the UAV-assisted network slicing problem through problem decomposition and mathematical methods, and takes into account SLA requirements, resulting in higher practical application value compared to traditional solutions.

[0104] This invention breaks the industry's technical prejudice that the problem of unmanned aerial vehicle (UAV) assisted network resource allocation cannot be solved quickly.

[0105] Third, the technical solution of the present invention solves the following technical problems of the prior art in industrial applications and achieves significant technical progress:

[0106] 1. Problems with existing technology:

[0107] Inefficient resource allocation: Existing UAV-assisted networks suffer from inefficiency and inaccuracy in resource allocation, failing to effectively respond to varying needs and environmental changes. Especially in the context of multi-channel signal processing and high-frequency requirements, traditional methods often cannot adjust resources according to the dynamic environment, leading to resource waste or poor signal quality.

[0108] Insufficient optimization of drone deployment: Traditional resource allocation methods often overlook the deployment efficiency and signal coverage of drones, which prevents drones from maximizing their communication assistance role in actual deployment and affects the overall network performance.

[0109] Complex optimization algorithms: Existing optimization algorithms are often complex and computationally intensive, resulting in slow processing speeds, especially in large-scale network or real-time communication environments, where they cannot meet real-time requirements.

[0110] 2. Technological advancements of this invention:

[0111] Effective Integration of Resource Allocation and Drone Deployment: This invention effectively improves the efficiency of resource allocation and the rationality of drone deployment by dividing the resource allocation problem and the drone deployment problem into two sub-problems and solving them through optimization algorithms. This solution can adjust resource configuration and drone deployment locations in real time, maximizing network signal quality and resource utilization, and solving the problem of unreasonable resource allocation in traditional technologies.

[0112] Implementation of the global optimal solution: By employing the KKT conditions and relaxation techniques for the pseudo-concave problem, this invention can quickly calculate the global optimal solution in complex multi-channel communication environments, ensuring the optimization of resource allocation and UAV deployment, and significantly improving network performance and stability.

[0113] Lightweight decomposition and optimal rounding algorithms: To improve computational efficiency, this invention designs lightweight decomposition and optimal rounding algorithms, effectively reducing computational complexity and rounding errors. This technical solution is particularly suitable for real-time data processing and resource scheduling in large-scale networks, enabling rapid response to changes and reduced latency in practical applications.

[0114] 3. Advantages of industrial applications:

[0115] Optimized signal coverage: This invention enables optimal deployment and resource allocation for drones in dynamic environments, effectively improving the signal quality and coverage of communication networks. It is particularly suitable for application scenarios requiring high signal strength and high bandwidth, such as drone communication, smart cities, and industrial IoT.

[0116] Improved system efficiency: By adopting the optimization method of this invention, network resources are maximized, avoiding the resource waste problem in traditional methods and improving the overall system efficiency.

[0117] Enhanced real-time processing capabilities: The optimized algorithm design of this invention enables the system to respond to and process real-time data quickly, greatly improving the system's real-time performance, adapting to the needs of dynamic network environments, and meeting the high standards of real-time performance and efficiency required in industrial applications.

[0118] In summary, the technical solution of this invention overcomes several technical bottlenecks in resource allocation and optimization algorithms based on existing technologies, significantly improves the performance of UAV-assisted networks, and has broad prospects for industrial applications. Attached Figure Description

[0119] Figure 1 This is a flowchart of a lightweight slice resource configuration method for UAV-assisted networks provided in an embodiment of the present invention.

[0120] Figure 2 This is a block diagram of a lightweight slice resource configuration system for UAV-assisted networks provided in an embodiment of the present invention.

[0121] Figure 3 This is a diagram illustrating the application scenario of UAWN slicing provided in an embodiment of the present invention.

[0122] Figure 4 This is a solution framework diagram provided by an embodiment of the present invention.

[0123] Figure 5 These are convergence time diagrams for different implementations provided in embodiments of the present invention.

[0124] Figure 6 This is a comparison chart of the maximum and minimum speeds of users under different network scenarios provided in the embodiments of the present invention.

[0125] (a) The impact of the number of users on the maximum-minimum user rate; (b) The impact of total bandwidth on the maximum-minimum user rate; (c) The impact of maximum power on the maximum-minimum user rate; (d) The algorithm convergence process and the change of the target value. Detailed Implementation

[0126] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0127] like Figure 1As shown, the lightweight slice resource configuration method for UAV-assisted networks provided in this embodiment of the invention includes the following steps:

[0128] S101, the problem is equivalently divided into a resource allocation problem and a drone deployment problem;

[0129] S102, by analyzing the properties of the subproblems, the explicit solutions to the subproblems are derived, and the explicit expression of the main problem is derived.

[0130] S103 proves that the relaxation principal problem is a pseudo-concave problem, and its stationary point corresponds to the global optimal solution. Based on this, a lightweight decomposition algorithm is used to find the global optimal solution using the KKT conditions. An optimal rounding algorithm is designed to minimize the rounding loss of the relaxation solution.

[0131] 1. Signal and data processing for resource allocation problems

[0132] In step 1, after dividing the problem into an equivalent resource allocation problem and a drone deployment problem, the resources in the network (such as spectrum, bandwidth, and computing power) are first precisely quantified. Signal data processing at this stage mainly involves mapping the spectrum allocation of the wireless communication network to the actual deployment requirements of the drones. Using an optimization model based on signal power control and spectrum utilization, the requirements of each communication link (e.g., bandwidth, latency, throughput) are compared with the current resource configuration to obtain a signal data allocation strategy. At this point, data acquisition and analysis rely on real-time channel information, drone location and mobility, and signal processing algorithms are used to estimate the interference and signal strength of each link, forming a preliminary resource allocation plan.

[0133] 2. Subproblem Analysis and Solution Derivation

[0134] In step 2, after analyzing the nature of the resource allocation problem and the UAV deployment problem, their explicit solutions are derived. At this stage, signal data processing requires relating the solutions to each sub-problem through calculations of signal interference and network topology. Specifically, the resource allocation sub-problem requires optimal allocation of bandwidth and computing resources in the UAV-assisted network, considering interference from different frequency channels, user signal reception strength, and power control. This process involves the application of signal enhancement and noise reduction techniques, optimizing the solutions to each sub-problem using digital signal processing algorithms. Simultaneously, the solution derivation process incorporates the coverage of the wireless signal and the adaptability of the spatiotemporal signal to derive the optimal configuration for the overall network.

[0135] 3. Relaxation of the pseudoconcave problem and calculation of the global optimal solution

[0136] In step 3, after the relaxation master problem is determined to be a pseudo-concave problem, signal data processing enters the optimization phase. In this phase, by analyzing the structural characteristics of the problem, the explicit expression of the relaxation problem is used to calculate stationary points, which correspond to the global optimum of the system. The key to signal data processing is using Karush-Kuhn-Tucker Conditions (KKT Conditions) to find the global optimum. Through the optimization relationship between constraints and the objective function, the signal data is input into the KKT condition solver to obtain the optimal resource allocation and UAV deployment scheme under the current network configuration. During this process, the signal data undergoes multiple iterations and adjustments to ensure that the signal strength and transmission quality of each wireless link are maximized.

[0137] 4. Applications of lightweight decomposition algorithms and optimal rounding algorithms

[0138] Finally, the signal data is refined using lightweight decomposition and optimal rounding algorithms. The rounding loss of the relaxed solution is minimized to ensure the application of the global optimal solution in practical implementation. Signal data processing at this stage involves mapping the calculated theoretical optimal value to practical applications, using an approximation algorithm to discretize resource allocation. The designed optimal rounding algorithm resolves the rounding error caused by discretization, ensuring that the resource allocation for each signal link does not suffer excessive performance loss in actual deployment. Throughout this process, interference control and reasonable spectrum allocation are fully guaranteed, ultimately achieving the optimal resource allocation scheme in the UAV-assisted network.

[0139] The derivation method provided in this embodiment of the invention is as follows:

[0140] Basis network model

[0141] First, let's focus on downlink transmission, where the drone is equipped with a maximum transmission power of P. max The antenna, the bandwidth available to the drone is denoted as B. tot Assuming the drone's projected position on the horizontal plane is the center of the drone network; and letting θ be the beamwidth and h be the drone's height; then the following coverage conditions must be met:

[0142] h·tanθ≥R (1)

[0143] In the above formula, R is the radius of the UAWN; considering that the flight capability of the UAWN is limited, its altitude is restricted by the following conditions:

[0144]

[0145] In the above formula, c represents the flight speed of the drone. T represents the drone's flight altitude in the previous time slot. sThe duration of each time slot in the network;

[0146] Network slicing model

[0147] use Let i represent the set of network slices, and let i be the set of user devices. u ij Let u represent the j-th user device in slice i. ij The two-dimensional coordinates are as follows User equipment u ij The square distance between the drone and the drone is in Let g0 be the channel power gain at a reference distance of one meter; according to existing literature, user equipment u ij The channel power gain between the drone and the other device can be modeled as follows:

[0148]

[0149] Where α is the path loss exponent; in This represents the small-scale fading channel coefficients of the Rayleigh distribution, that is... Follows a non-central chi-square distribution;

[0150] Use B i Let x represent the channel bandwidth of slice i; it should be noted that due to service differentiation, the channel bandwidth of different slices is different; use a non-negative integer x. ij Indicates allocation to user equipment u ij The number of channels; obviously, x ij The following capacity constraints must be met:

[0151]

[0152] Among them, drones deliver to user equipment u ij The transmitted power is denoted as p. ij The following capacity constraints must be met:

[0153]

[0154] To guarantee the SLA, each slice should provide a guaranteed data rate to subscribers, i.e.:

[0155]

[0156] Where, σ 2 It is the power of the background noise. This is the minimum rate guaranteed by slice i; in addition, each user equipment can request a customized service rate. Considering the tractability of the problem, the constraint is modeled as follows:

[0157]

[0158] in, It is a positive integer and satisfies

[0159] UAWN network slicing model

[0160] The UAWN network slicing problem can be described as follows:

[0161]

[0162] st(1)-(2),(4)-(7)(9)

[0163] h min ≤h≤h max (10)

[0164]

[0165] Among them, h min h is the minimum permissible flight altitude for drones. max The problem equivalently divided according to the maximum permissible flight altitude of the drone is as follows:

[0166] USP decomposition

[0167] The objective function of USP is denoted as f(x,p,h) = min i,j f ij (x ij ,p ij The so-called problem partitioning refers to projecting both the objective function and constraint set of the USP onto the (x,p) axis. After projecting the USP onto the (x,p) axis, the following main problem can be obtained:

[0168]

[0169] st(x,p)∈V(14)

[0170] Where X and P are defined as follows: X:={x|x satisfies (4) and (7)}, and P:={p|p satisfies (5) and (11)}; set V is defined as:

[0171] V: = {p | there exists h∈H} B Make p satisfy (6)} (15)

[0172] In the above formula, H B It is the interval set defined by constraints (1), (2) and (10), that is in and They are respectively and

[0173] In MP, v(x,p) is defined as a subproblem:

[0174]

[0175] Explicit solutions to subproblems

[0176] The explicit solution to SP can be directly derived. This solution is independent of x and p, and the following theorem about SP can be derived:

[0177] For each fixed pair If the corresponding SP is feasible, then its optimal solution must be

[0178] Explicit expression of the main problem and its relaxation

[0179] Let n represent the index pair (i,j). This refers to all user equipment in the network, i.e. The number of user devices in the system is represented by N. The relaxed MP can be represented as the following problem:

[0180]

[0181] std T x≤B tot (19)

[0182] 1 T p≤P max (20)

[0183] x≥ x ,p≥ p (twenty one)

[0184] Where d and 1 are the coefficients of constraints (4) and (5), respectively. β n For n, the index pair (i,j) contains B i , To make it clearer, let ω(x,p) denote the objective function of MP-Relax; and it can be proven that the inequality constraints of constraints (19) and (20) can be replaced by equality constraints.

[0185] The lightweight decomposition algorithm for finding the global optimum using KKT conditions provided in this embodiment of the invention:

[0186] Solving the relaxation master problem

[0187] The BCD method will be used to solve MP-Relax. Based on pseudo-convex optimization theory and the characterization of MP-Relax solutions, two subproblems, namely the power allocation subproblem and the sub-channel allocation subproblem, will be solved using the KKT conditions. Then, a BCD algorithm for solving the global optimal solution of MP-Relax will be proposed.

[0188] Power allocator problem

[0189] By fixing a stationary point x on the domain, we can obtain the following power allocation subproblem:

[0190]

[0191] st1 T p≤P max (twenty three)

[0192] P≥ P (twenty four)

[0193] Among them, a n =β n x n It is evident that the power allocator problem is a non-smooth convex programming problem. Since its objective function is strictly convex, it has a unique solution; assuming this optimal solution is (p... * ,z * ),definition:

[0194]

[0195] Sub-channel allocation sub-problem;

[0196] The BCD algorithm for relaxing the master problem.

[0197] The sub-channel allocation sub-problem provided in this embodiment of the invention:

[0198] At the x-coordinate, the following sub-channel allocation sub-problem needs to be solved for a fixed p:

[0199]

[0200] std T x≤B max (28)

[0201] x≥ x (29)

[0202] Where c n =β n ln(1+b n p n ); due to c n ≠0, Sub x The objective function is strictly convex, which indicates the uniqueness of its optimal solution; similarly, assuming the optimal value is t * Assume the optimal solution is It can be deduced that:

[0203]

[0204] The BCD algorithm for the relaxed principal problem provided in this embodiment of the invention:

[0205] Based on KKT x and KKT p The algorithm proposes a BCD algorithm based on iterative rules to solve MP-Relax.

[0206] like Figure 2 As shown, an embodiment of the present invention provides a lightweight slice resource configuration system for UAV-assisted networks, comprising:

[0207] The partitioning module is used to divide the problem into two equivalent problems: resource allocation and drone deployment.

[0208] The derivation module is used to derive explicit solutions to subproblems and explicit expressions of the main problem by analyzing the properties of subproblems.

[0209] The solution module is used to prove that the relaxation master problem is a pseudo-concave problem, and its stationary point corresponds to the global optimum. Based on this, a lightweight decomposition algorithm is used to find the global optimum using the KKT conditions, and an optimal rounding algorithm is designed to minimize the rounding loss of the relaxation solution.

[0210] Another object of the present invention is to provide a computer device including a memory and a processor, the memory storing a computer program, which, when executed by the processor, causes the processor to perform the steps of the lightweight slice resource allocation method for unmanned aerial vehicle-assisted networks.

[0211] Another object of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of the lightweight slice resource allocation method for unmanned aerial vehicle-assisted networks.

[0212] Another objective of this invention is to provide an information data processing terminal for implementing the lightweight slice resource configuration system for UAV-assisted networks.

[0213] Specific implementation of the present invention:

[0214] 1. System Model

[0215] Basis network model

[0216] like Figure 3 As shown, the focus is on a cellular drone-assisted wireless network (UAWN 1), consisting of a drone and several ground small base stations. Each cell comprises multiple user equipment (UEs) and a ground base station providing services to them. In the considered scenario, it is assumed that the base station cannot provide satisfactory QoS for the UEs, therefore the drone is deployed to supplement the base station. The designed framework operates in a time-slot manner, where the drone's altitude is periodically updated in each time slot to adapt to network conditions. The duration of each time slot is denoted as T. s .

[0217] In this invention, the focus is first on downlink transmission, wherein the UAV is equipped with a maximum transmission power of P. max The antenna, the bandwidth available to the drone is denoted as B. tot Assume the projected position of the drone on the horizontal plane is the center of the drone network. Let θ be the beamwidth and h be the drone's altitude. Then the following coverage conditions must be satisfied:

[0218] h·tanθ≥R (1)

[0219] In the above formula, R is the radius of the UAWN. Considering that the flight capability of a drone is limited, its altitude is restricted by the following conditions:

[0220]

[0221] In the above formula, c represents the flight speed of the drone. This represents the drone's flight altitude in the previous time slot.

[0222] Network slicing model

[0223] Due to the heterogeneity of user equipment, the underlying network can be divided into several network slices, each of which will be customized to support a specific type of service. Let i represent the set of network slices, and let i be the set of user devices. u ij Let u represent the j-th user device in slice i. ij The two-dimensional coordinates are as follows User equipment u ij The square distance between the drone and the drone is in Let g0 be the channel power gain at a reference distance of one meter. Based on existing literature, the user equipment u... ij The channel power gain between the drone and the other device can be modeled as follows:

[0224]

[0225] Where α is the path loss exponent. in This represents the small-scale fading channel coefficients of the Rayleigh distribution, that is... It follows a non-central chi-square distribution.

[0226] Use B i This represents the channel bandwidth of slice i. It's important to note that due to service differentiation, the channel bandwidth differs between different slices. Let x be a non-negative integer. ij Indicates allocation to user equipment u ij The number of channels. Clearly, x ij The following capacity constraints must be met:

[0227]

[0228] Among them, drones deliver to user equipment u ij The transmitted power is denoted as p. ij The following capacity constraints must be met:

[0229]

[0230] 6G end-to-end network slicing spans multiple domains and adjusts multi-dimensional service level agreements (SLAs) in terms of latency, throughput, and reliability. Considering that the main part of the UAWN is the Radio Access Network (RAN), the model uses transmission rate as a key SLA metric. To guarantee the SLA, each slice should provide a guaranteed data rate to subscribers, i.e.:

[0231]

[0232] Where, σ 2 It is the power of the background noise. This is the minimum rate guaranteed by slice i. Furthermore, each user equipment can request a customized service rate. Considering the tractability of the problem, the constraint is modeled as follows:

[0233]

[0234] in, It is a positive integer and satisfies

[0235] UAWN network slicing model

[0236] In the UAWN Network Slicing Problem, my goal is to find the optimal values ​​of channel allocation x, power allocation p, and UAV altitude h, while maximizing the minimum user rate. The UAWN Network Slicing Problem (USP) can be formulated as follows:

[0237]

[0238] st(1)-(2),(4)-(7)(9)

[0239] h min ≤h≤h max (10)

[0240]

[0241] In USP, constraint (10) ensures that the drone's altitude is limited by a lower limit h. min It is constrained to avoid collisions with ground obstacles, and is also subject to an upper limit h. max The constraint is the maximum allowed flight altitude of the UAV. It can be seen that USP is a large-scale non-convex mixed integer problem. Even if x is relaxed to a continuous variable, the objective function (8) is still non-smooth and non-convex. In addition, USP contains a large number of non-convex constraints as shown in equation (6), which makes solving USP extremely difficult.

[0242] 2. Problem decomposition and drone deployment optimization

[0243] This chapter will explain how to solve the BCD of the USP. 2 The framework, its main steps are as follows Figure 4 As shown, the framework mainly consists of five steps. First, the USP is divided into a main problem and subproblems using projection techniques. Second, the structure of the subproblems is examined and their closed-form solutions are derived, obtaining the optimal solution h for each fixed (x,p). * Therefore, the explicit expression of the main problem is derived. The third step is to relax the integer constraints of the main problem to obtain MP-Relax, proving that it is a pseudo-concave problem and that each stationary point corresponds to a global optimum. The fourth step is to find the global optimum of MP-Relax using batch coordinate descent: decomposing it into two subproblems and solving them iteratively in a cyclic order; by analyzing the KKT conditions of these two subproblems, the scheme designs two low-complexity algorithms to solve them, thereby accelerating BCD. 2 The final step proposes an Optimal Rounding Algorithm (ORA) to round the relaxed solution.

[0244] USP decomposition

[0245] The objective function of USP is denoted as f(x,p,h) = min i,j f ij (x ij ,p ij The so-called problem partitioning refers to projecting both the objective function and constraint set of the USP onto the (x,p) axis. After projecting the USP onto the (x,p) axis, the following main problem can be obtained:

[0246]

[0247] st(x,p)∈V(14)

[0248] Where X and P are defined as follows: X:={x|x satisfies (4) and (7)}, and P:={p|p satisfies (5) and (11)}. The set V is defined as:

[0249] V: = {p | there exists h∈H} B Make p satisfy (6)} (15)

[0250] In the above formula, H B It is the interval set defined by constraints (1), (2) and (10), that is in and They are respectively and

[0251] In MP, v(x,p) is defined as a subproblem:

[0252]

[0253] Explicit solutions to subproblems

[0254] The explicit solution to SP can be directly derived. This solution is independent of x and p, and the following theorem about SP can be derived:

[0255] For each fixed pair If the corresponding SP is feasible, then its optimal solution must be

[0256] Explicit expression of the main problem and its relaxation

[0257] Substitute h into h *MP can be equivalently transformed into a joint channel and power allocation problem. In this case, MP is a non-smooth and non-convex mixed integer problem, making it extremely difficult to solve. This invention does not directly address it, but first solves its relaxed version (i.e., MP-Relax) to obtain a relaxed optimal solution. Then, an optimal rounding algorithm is used to round the relaxed solution to an integer value. For simplified notation, n represents the index pair (i,j). This refers to all user equipment in the network, i.e. The number of user devices in the system is represented by N. The relaxed MP can be represented as the following problem:

[0258]

[0259] std T x≤B tot (19)

[0260] 1 T p≤P max (20)

[0261] x≥ x ,p≥ p (twenty one)

[0262] Where d and 1 are the coefficients of constraints (4) and (5), respectively. β n For n, the index pair (i,j) contains B i , To make it clearer, let ω(x,p) denote the objective function of MP-Relax. It can also be proven that the inequality constraints in (19) and (20) can be replaced by equality constraints.

[0263] 3. Solving the master relaxation problem

[0264] The BCD method will be used to solve MP-Relax. Based on pseudo-convex optimization theory and the characterization of MP-Relax solutions, two subproblems are solved using the KKT conditions: the power allocation subproblem and the sub-channel allocation subproblem. Then, a BCD algorithm for solving the global optimal solution of MP-Relax is proposed.

[0265] Power allocator problem

[0266] By fixing a stationary point x on the domain, we can obtain the following power allocation subproblem:

[0267]

[0268] st1 T p≤P max (twenty three)

[0269] P≥ P (twenty four)

[0270] Among them, a n =β n x n It is evident that the power allocator problem is a non-smooth convex programming problem. Since its objective function is strictly convex, it has a unique solution. Assume this optimal solution is (p... * ,z * ),definition:

[0271]

[0272] To find this optimal solution, a power allocation algorithm based on KKT is proposed, or KKT for short. p It is used to solve the power distribution subproblem.

[0273]

[0274] Sub-channel allocation sub-problem

[0275] At the x-coordinate, the following sub-channel allocation sub-problem needs to be solved for a fixed p:

[0276]

[0277] std T x≤B max (28)

[0278] x≥ x (29)

[0279] Where c n =β n ln(1+b n p n Due to c n ≠0, Sub x The objective function is strictly convex, which indicates the uniqueness of its optimal solution. Similarly, assume the optimal value is t. * Assume the optimal solution is It can be deduced that:

[0280]

[0281] To find the optimal value, a subchannel allocation algorithm based on KKT is proposed, abbreviated as KKT. x It is used to solve the subchannel allocation subproblem.

[0282]

[0283] BCD algorithm for relaxing the master problem

[0284] Based on KKT x and KKT p The following BCD algorithm is proposed for solving MP-Relax.

[0285]

[0286]

[0287] 4. Simulation Results

[0288] Figure 5 Convergence time for different implementations.

[0289] Figure 6 Comparison of maximum and minimum user speeds under different network scenarios.

[0290] This invention addresses the slicing problem in UAV-assisted networks by proposing a low-complexity optimization framework to jointly optimize resource allocation and UAV deployment. The problem is equivalently divided into a resource allocation problem and a UAV deployment problem. By analyzing the properties of the subproblems, explicit solutions to the subproblems are derived, leading to an explicit expression of the master problem. It is then proven that the relaxed master problem is a pseudo-concave problem, with its stationary point corresponding to the global optimum. Based on this, a lightweight decomposition algorithm using KKT conditions to find the global optimum is proposed. Furthermore, an optimal rounding algorithm is designed to minimize the rounding loss of the relaxed solution. Numerical results demonstrate that, compared to existing algorithms, the proposed algorithm has significant advantages such as faster convergence speed and higher user throughput. Figure 6 The convergence curves of the algorithm in this invention and existing algorithms are shown. USP is a large-scale non-convex mixed integer programming problem, and solutions are lacking in existing research both domestically and internationally. Considering that metaheuristics can evade local optima in non-convex problems, this invention selects the following three algorithms as comparison benchmarks: Quasi-subgradient Projection Algorithm (QPA); Adaptive Inertia Weighted Particle Swarm Optimization (AIW-PSO); and Success History Intelligent Optimizer (SHIO). The convergence curves of BCD are also described. 2 Compared with the convergence curves of the three existing algorithms, this invention sets the number of UEs in UAWN to 60, and it can be seen that BCD 2 Convergence is achieved in just two iterations, equivalent to one round of BCD iteration. In contrast, QPA requires approximately 100 iterations to converge, while AIW-PSO and SHIO fail to converge at all. Therefore, this invention concludes that the BCD2 of this invention is lightweight and converges quickly.

[0291] Comparison of the algorithm of this invention with the QPA algorithm.

[0292]

[0293]

[0294] To quantify the convergence speed of BCD, the table above compares the convergence speeds of BCD. 2 The execution time and optimal value of QPA vary with the number of UEs, ranging from 20 to 100. BCD can be observed. 2 Convergence occurs within milliseconds, which is approximately 0.02% of QPA. Furthermore, BCD can be observed. 2 The optimal value is slightly greater than QPA. This result indicates that BCD 2 In a sense, it is an exact algorithm that can find the exact solution in a finite number of iterations, which QPA cannot achieve. Furthermore, the convergence time increases linearly with the problem size, verifying the log-linear convergence of BCD2.

[0295] It should be noted that embodiments of the present invention can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of the above-described hardware circuitry and software, such as firmware.

[0296] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A lightweight slicing resource allocation method for unmanned aerial vehicle (UAV) assisted networks, characterized in that, Includes the following steps: Step 1: Establish the UAWN network slicing model and the UAWN network slicing problem. Use the projection method to decompose the UAWN network slicing problem into a main problem and sub-problems and solve them separately. The main problem is equivalently divided into a power allocation sub-problem and a sub-channel allocation sub-problem. By quantifying the resources in the network and using signal processing algorithms to estimate the interference and signal strength of each link, a preliminary resource allocation scheme is formed. Step 2: In solving the main problem, the main problem is relaxed to MP-Relax, and the global optimal solution of MP-Relax is found using the Batch Coordinate Descent (BCD) method. This is decomposed into two subproblems: a power allocation subproblem and a sub-channel allocation subproblem, which are solved iteratively in a cyclical order. Based on pseudo-convex optimization theory and the representation of MP-Relax solutions, the two subproblems are solved using the KKT conditions. The optimal rounding algorithm ORA is used to round the relaxed solution. Finally, signal data is calculated using optimization relationships to ensure optimal resource allocation and UAV deployment schemes under network configuration. Step 3: Analyze the nature of the resource allocation problem and the UAV deployment problem, derive their respective explicit solutions, and optimize the solution of each sub-problem through signal enhancement, noise reduction technology and digital signal processing algorithms, taking into account the coverage of wireless signals and spatiotemporal signal adaptability; Step 4: The signal data is refined by using a lightweight decomposition algorithm and an optimal rounding algorithm to minimize the rounding loss of the relaxed solution, ensure the practical application of the global optimal solution, solve the error caused by the discretization operation, and optimize the allocation of signal link resources and interference control.

2. The lightweight slice resource allocation method for UAV-assisted networks as described in claim 1, characterized in that, The resource allocation plan is as follows: Basis network model First, let's focus on downlink transmission, where the drone is equipped with a maximum transmission power of The antenna, the bandwidth available to the drone is denoted as Assume the projected position of the drone on the horizontal plane is the center of the drone network; let... For beamwidth, The altitude of the drone must be specified; therefore, the following coverage conditions must be met: (1) In the above formula Let UWN be the radius; considering the limited flight capabilities of drones, its altitude is restricted by the following conditions: (2) In the above formula For the flight speed of the drone, The drone's flight altitude in the previous time slot. The duration of each time slot in the network; Network slicing model use Represents a set of network slices, slices The set of user equipment in the data is denoted as , Slice The first in User equipment, user equipment The two-dimensional coordinates are as follows User equipment The square distance between the drone and the drone is ,in ;set up Channel power gain at a reference distance of one meter; User equipment The channel power gain model between the drone and the other device is as follows: (3) in, It is the path loss index; ,in This represents the small-scale fading channel coefficients of the Rayleigh distribution, that is... Follows a non-central chi-square distribution; use Slice The channel bandwidth; it should be noted that due to service differentiation, the channel bandwidth differs between different slices; use non-negative integers. Indicates allocation to user equipment The number of channels; obviously, The following capacity constraints must be met: (4) Among them, drones deliver to user equipment The transmitted power is denoted as The following capacity constraints must be met: (5) To guarantee the SLA, each slice should provide a guaranteed data rate to subscribers, i.e.: (6) in, It is the power of the background noise. It is a slice The guaranteed minimum rate; in addition, each user device requests a customized service rate. Considering the tractability of the problem, this constraint is modeled as follows: (7) in, It is a positive integer and satisfies ; UAWN network slicing model The UAWN network slicing problem is described as follows: (8) (9) (10) (11) (12)； in, This refers to the minimum permissible flight altitude for drones. This is the maximum permissible flight altitude for drones.

3. The lightweight slice resource allocation method for UAV-assisted networks as described in claim 2, characterized in that, The UAWN network slicing problem is equivalently divided into: USP decomposition The scheme denotes the objective function of USP as follows: The so-called problem partitioning refers to simultaneously projecting the USP's objective function and constraint set onto... On the axis, project the USP onto The following main problem arises after the axis is aligned: (13) (14) in, and They are defined as follows: ,as well as ;gather Defined as: (15) In the above formula It is the interval set defined by constraints (1), (2) and (10), that is ,in and They are respectively and ; In MP, Defined as a subproblem: (16) (17) Explicit solutions to subproblems The explicit solution to SP was directly derived, and this solution is consistent with... and This is irrelevant, and the following theorem about SP is derived: For each fixed pair If the corresponding SP is feasible, then its optimal solution must be... ; Explicit expression of the main problem and its relaxation use Indicates index pairs , This refers to all user equipment in the network, i.e. The number of user devices in the system is used The relaxed MP can be represented by the following problem: (18) (19) (20) (21) in, and These are the coefficients of constraints (4) and (5), respectively. , , , , For the index pair corresponding to n Below , , , , for Corresponding index pairs In , To make it clearer, we will use... Let represent the objective function of MP-Relax; and prove that the inequality constraints of constraints (19) and (20) are replaced by equality constraints.

4. The lightweight slice resource allocation method for UAV-assisted networks as described in claim 3, characterized in that, A lightweight decomposition algorithm for finding the global optimum using KKT conditions: Solving the relaxation master problem The BCD method will be used to solve MP-Relax. Based on pseudo-convex optimization theory and the characterization of MP-Relax solutions, two subproblems, namely the power allocation subproblem and the sub-channel allocation subproblem, will be solved using the KKT conditions. Then, a BCD algorithm for solving the global optimal solution of MP-Relax will be proposed. Power allocator problem By fixing a stationary point on the domain This leads to the following power allocation subproblem: (22) (23) (24) in, It is evident that the power allocator problem is a non-smooth convex programming problem. Since its objective function is strictly convex, it has a unique solution; let this optimal solution be... ,definition: (25) (26) Sub-channel allocation sub-problem exist At the coordinates, it is necessary to consider a fixed... Solve the following sub-channel allocation sub-problem: (27) (28) (29) in ;because , The objective function is strictly convex, which indicates the uniqueness of its optimal solution; similarly, assuming the optimal value is... Assume the optimal solution is It can be deduced that: (30) (31) BCD algorithm for relaxing the master problem based on and The algorithm proposes a BCD algorithm based on iterative rules to solve MP-Relax.

5. A lightweight slice resource allocation system for UAV-assisted networks, implementing the lightweight slice resource allocation method for UAV-assisted networks as described in any one of claims 1-4, characterized in that, A lightweight slicing resource allocation system for UAV-assisted networks includes: The partitioning module is used to divide the problem into two equivalent problems: resource allocation and drone deployment. The derivation module is used to derive explicit solutions to subproblems and explicit expressions of the main problem by analyzing the properties of subproblems. The solution module is used to prove that the relaxation master problem is a pseudo-concave problem, and its stationary point corresponds to the global optimum. Based on this, a lightweight decomposition algorithm is used to find the global optimum using the KKT conditions, and an optimal rounding algorithm is designed to minimize the rounding loss of the relaxation solution.

6. A computer device, characterized in that, The computer device includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the lightweight slice resource allocation method for UAV-assisted networks as claimed in any one of claims 1-4.

7. A computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the lightweight slice resource allocation method for unmanned aerial vehicle-assisted networks as claimed in any one of claims 1-4.

Images