Joint deployment and resource optimization method based on multi-unmanned aerial vehicle assisted wireless network

Abstract

The application discloses a kind of based on the joint deployment and resource optimization method of multi-unmanned aerial vehicle assisted wireless network, comprising: establishing the communication system model based on multi-unmanned aerial vehicle assisted wireless network;Design optimization target, establish optimization problem;The optimization target is to minimize system cost;The minimization system cost problem is decomposed into two sub-optimization problems, i.e.: unmanned aerial vehicle position deployment optimization problem and unmanned aerial vehicle setting optimization problem;The unmanned aerial vehicle setting includes: unmanned aerial vehicle power, communication bandwidth, user and unmanned aerial vehicle association and unmanned aerial vehicle quantity;Given unmanned aerial vehicle position deployment, solve unmanned aerial vehicle setting optimization problem;Given unmanned aerial vehicle setting, solve unmanned aerial vehicle position deployment optimization problem;According to the solution obtained above, the minimization system cost problem is completed, i.e. based on the joint deployment and resource optimization of multi-unmanned aerial vehicle assisted wireless network.

Description

Technical Field

[0001] This invention relates to a method for the joint deployment and resource optimization of wireless networks, and more particularly to a method for the joint deployment and resource optimization of multi-UAV-assisted wireless networks. Background Technology

[0002] Unmanned aerial vehicles (UAVs) have gained significant attention in recent years due to their flexible deployment and relatively low cost. UAVs have diverse applications across various sectors, including telecommunications, rescue operations, aerial sensing for security purposes, surveillance, package delivery, and precision agriculture. In future wireless networks, UAVs are expected to play a crucial role in enhancing the communication capacity and coverage of terrestrial networks. Furthermore, the line-of-sight (LoS) air-to-ground (A2G) channel between user equipment (UEs) and UAVs can mitigate signal congestion. In remote areas, sparsely deployed ground base stations (GBS) alone often cannot meet communication needs. In such cases, cost-effective UAV-assisted wireless networks are a more practical choice. In disaster areas, UAVs are also crucial in providing emergency connectivity for wireless devices, as traditional cellular wireless coverage is largely unavailable.

[0003] UAVs can serve as static aerial base stations, hovering at selected locations and providing wireless services to users. Compared to traditional base stations (GBS), deploying multiple UAVs to assist wireless communication can achieve higher throughput. Various studies have explored the optimal three-dimensional deployment of UAVs, including maximizing ground user coverage, average rate at the worst-case bit error rate threshold, and energy efficiency. These methods have application value for enhancing wireless coverage and capacity. However, deploying a large number of UAVs to provide comprehensive wireless services can impose significant investment costs on network providers. Furthermore, the capacity of the wireless backhaul link between UAVs and GBS is a key factor limiting system performance, which also depends on the number of UAVs deployed. Therefore, in multi-UAV-assisted wireless networks, there is a trade-off between system throughput and network deployment costs. Based on the advantages and characteristics of UAV communication systems mentioned above, UAVs as static aerial base stations providing wireless communication services have been extensively studied.

[0004] Sun and Masouros, in their paper "Deployment strategies of multiple aerial base stations for user coverage and power efficiency maximization," proposed a simple continuous deployment method and a clustering-based approach to maximize user coverage and power efficiency. Furthermore, Wang et al., in their paper "Deployment and association of multiple UAVs in UAV-assisted cellular networks with the knowledge of statistical user position," proposed a centralized multi-agent learning algorithm for UAV deployment and association schemes in UAV-assisted cellular networks to minimize power consumption for both users and UAVs during uplink transmission. However, these studies primarily focus on deploying a fixed number of UAVs. In multi-UAV-assisted wireless networks, how to consider the number of UAVs required for system deployment is crucial for the effective utilization of hardware resources.

[0005] A key challenge in UAV-assisted communication networks is the wireless backhaul link. In practical UAV wireless networks, UAVs typically connect to the GBS via a wireless backhaul link, which can be enhanced using millimeter-wave frequencies. However, the capacity of the wireless backhaul link is highly sensitive to the location and propagation environment of the UAVs. Furthermore, the number of deployed UAVs also affects the backhaul link capacity, making it a critical limiting factor for system performance. In their 2020 paper, "Multiple uav-mounted base station placement and user association with joint fronthaul and backhaul optimization," C. Qiu et al. transformed the system optimization problem with backhaul capacity constraints into an unconstrained problem and used gradient descent to determine the UAV's location.

[0006] In summary, the problems with existing technologies are:

[0007] (1) In a multi-UAV communication system, due to the existence of co-channel interference, other factors need to be considered to optimize the UAV's location selection.

[0008] (2) In a multi-UAV assisted wireless network, how to consider the number of UAVs required for system deployment is crucial for the effective use of hardware resources.

[0009] (3) In practical applications, the selection of UAV locations needs to be considered more carefully due to the limited capacity of the backhaul link. The capacity limitation of the backhaul link will affect the deployment and communication capabilities of the UAV.

[0010] (4) Therefore, when designing a drone communication system, it is necessary to comprehensively consider multiple factors such as user geographical location, co-channel interference, backhaul link capacity and drone deployment cost, and adopt appropriate optimization methods to deal with these problems in order to obtain the best system performance and user experience. Summary of the Invention

[0011] Purpose of the invention: The technical problem to be solved by the present invention is to provide a method for joint deployment and resource optimization based on multi-UAV assisted wireless networks, which addresses the shortcomings of the existing technology.

[0012] To address the aforementioned technical problems, this invention discloses a method for joint deployment and resource optimization based on a multi-UAV-assisted wireless network, comprising the following steps:

[0013] Step 1: Establish a communication system model based on a multi-UAV assisted wireless network;

[0014] Step 2: For the communication system model, design optimization objectives and establish optimization problems;

[0015] Step 3: Decompose the problem of minimizing system cost in Step 2 into two sub-optimization problems: the drone location deployment optimization problem and the drone setup optimization problem.

[0016] Step 4: Given the deployment locations of the drones, solve the drone setup optimization problem;

[0017] Step 5: Given the drone settings, solve the drone location deployment optimization problem;

[0018] Step 6: Based on the solutions obtained in Steps 4 and 5, complete the problem of minimizing system cost, that is, realize the joint deployment and resource optimization based on multi-UAV assisted wireless network.

[0019] Furthermore, the optimization objective described in step 2 is to minimize the system cost, and the optimization problem is the system cost minimization problem.

[0020] Furthermore, the drone settings described in step 3 include: drone power, communication bandwidth, user-drone association, and number of drones.

[0021] Furthermore, the communication system model described in step 1 is as follows:

[0022] A multi-antenna ground base station provides communication services within a region via a group of drones; it employs a time-division multiple access downlink communication link to combine the drones with user communication equipment; the wireless communication system consists of a ground base station, J drones, and K user communication equipment, specifically including:

[0023] The collection of drones in the communication system is There are J drones, labeled {1,2,…,J}; the set of user communication devices is... There are K devices in total, labeled {1,2,…,K}; a communication connection needs to be established between UAV j and user equipment k, labeled x. j,k =1, otherwise marked as x j,k =0; a j Indicates the working status of the j-th drone, whether it is turned on or off, α jk This represents the communication bandwidth between the j-th drone and the k-th user device.

[0024] Furthermore, the optimization objective described in step 2 is to minimize the system cost, and the communication connection relationship x between the UAV and the user equipment is determined based on the optimization objective. j,k Communication bandwidth allocation between UAVs and user equipment α j,k 3D hovering position of the drone The transmit power p of drones as airborne base stations j ;

[0025] The problem of minimizing system cost is defined as an optimization model P1, as follows:

[0026]

[0027]

[0028]

[0029]

[0030] C4:,α jk ≤x jk

[0031] C5:,x jk ≤a j

[0032]

[0033]

[0034]

[0035]

[0036]

[0037]

[0038] Where, min indicates that the optimization objective is to minimize, st indicates that the subsequent expression represents the constraints, and the subscripts j and k represent the j-th UAV and the k-th user equipment, respectively; r k Let k represent the communication rate of user equipment k, and the optimization objective is to minimize the system cost.

[0039] In constraint C1 This represents the communication quality requirements of user equipment k;

[0040] In constraint C2, C j (d 0,j ) represents the upper limit of the backhaul link capacity between the UAV j and the ground base station, where d 0,j Let j be the physical distance between the drone and the ground base station;

[0041] In constraint C3 This represents the transmit power threshold of drone j;

[0042] Constraint C4 represents the allocated bandwidth fraction α when the k-th user and the j-th drone are not connected. jk =0;

[0043] Constraint C5 indicates that the drone providing the service is active;

[0044] Constraint C6 means that each user is served by only one drone;

[0045] Constraint C7 states that only active drones can connect to the user, and the communication bandwidth α for each drone is... jk The sum is less than 1;

[0046] Constraints C8 and C9 represent a j and x jk It is binary;

[0047] Optimization variables in constraints C10-C12 This represents the three-dimensional hovering position vector of the UAV, where the components are... These represent the coordinate values ​​on the X, Y, and Z axes in a three-dimensional coordinate system; in the constraints and This represents the lower and upper bounds of the coordinate vector, used to limit the geographical range of where the drone can hover.

[0048] Furthermore, the drone setup optimization problem described in step 3, namely, given the drone's location, yields the sub-problem P1.1 of optimizing drone power, communication bandwidth, user-drone association, and the number of drones, as detailed below:

[0049]

[0050] stP1.C1-P1.C9

[0051] The constraints of subproblem P1.1 are constraints C1-C9 in optimization model P1, where the position of the UAV is known.

[0052] Furthermore, the drone location deployment optimization problem described in step 3, that is, given the drone power, communication bandwidth, user-drone association, and number of drones, yields the sub-problem P1.2 for optimizing drone location, as detailed below:

[0053]

[0054]

[0055] P1.C2-P1.C3

[0056] P1.C10-P1.C12

[0057] The constraints of subproblem P1.2 are constraints C2-C3 and C10-C12 in optimization model P1, where constraint C1... This represents the lower bound of the rate of user equipment k, and the drone power, communication bandwidth, user-drone association, and number of drones are known.

[0058] Furthermore, the specific method for solving the UAV setup optimization sub-problem in step 4 includes:

[0059] For the communication connection variable x j,k and the drone switch state variable a j Introduce a penalty term into the objective function: as well as This makes the above optimization variables converge to the set of binary variables {0,1}, where λ in the penalty term is the penalty coefficient;

[0060] For other non-convex constraints in subproblem P1.1, a continuous convex approximation algorithm is adopted, using a first-order Taylor expansion as a substitute function for the non-convex function to transform the non-convex constraints into convex constraints. This formulation is further transformed into iteratively solving a series of standard convex optimization problems to obtain the solution to subproblem P1.1.

[0061] Furthermore, the specific method for solving the UAV location deployment optimization problem in step 5 includes:

[0062] For a series of non-convex constraints in subproblem P1.2, a continuous convex approximation algorithm is adopted, using a first-order Taylor expansion as a substitute function for the non-convex function to transform the non-convex constraints into convex constraints. This formulation is further transformed into iteratively solving a series of standard convex optimization problems, thereby obtaining the solution to subproblem P1.2.

[0063] Furthermore, step 6 involves solving the problem of minimizing system cost by using the block coordinate descent method to perform alternating iterative optimization on the results obtained in steps 4 and 5 to obtain the minimum system cost.

[0064] Beneficial effects:

[0065] 1. Improve communication quality: Using drones as relays can significantly improve the communication quality of user communication devices. Through their flexible mobility, drones can optimize communication transmission paths and improve signal strength and anti-interference capabilities.

[0066] 2. Reduced construction costs: Drone relay networks can reduce the overall construction cost of a communication system. Compared to traditional infrastructure construction, drone relay networks are more flexible in deployment and do not require large-scale infrastructure construction, thus saving costs.

[0067] 3. Establishment of Temporary Communication Networks: Drone relay networks provide a guarantee for the establishment of temporary communication networks. In scenarios such as disaster relief, emergency communications, or temporary events, drone relay networks can be deployed quickly to fill communication coverage gaps and provide reliable communication services.

[0068] 4. Consideration of Backhaul Capacity Limitations and Co-channel Interference Management: This invention considers the limited backhaul capacity when UAVs act as relay units and incorporates co-channel interference management into its scope. By optimizing the hovering position and deployment strategy of UAVs, the system throughput is further improved.

[0069] 5. Resource Saving and Flexible Deployment: By rationally planning the deployment number of drones, this invention saves communication resources and avoids unnecessary waste. At the same time, the flexible deployment of drones can adapt to different scenarios and needs, improving the flexibility and adaptability of the entire communication system. Attached Figure Description

[0070] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0071] Figure 1 This is a schematic diagram of the overall invention.

[0072] Figure 2 This is a schematic diagram of the overall process of the present invention.

[0073] Figure 3 This is a flowchart illustrating the process of iteratively solving subproblems to obtain the final optimized method.

[0074] Figure 4 This is a schematic diagram of the convergence result of the block coordinate descent algorithm.

[0075] Figure 5a This is a two-dimensional representation of user association and drone location using the block coordinate descent algorithm.

[0076] Figure 5b This is a 3D representation diagram of the algorithm used to associate users and the location of drones.

[0077] Figure 6 This is a schematic diagram of drone locations used to determine user associations using the MeanShift clustering method.

[0078] Figure 7 This is a diagram illustrating the deployment costs and power consumption of drones using different deployment methods. Detailed Implementation

[0079] This invention discloses a communication system optimization method based on multi-UAV assisted communication. For example... Figure 1 As shown, this method aims to deploy multiple drones to provide downlink communication services to multiple user devices with communication quality requirements within a target area, under conditions of limited drone relay backhaul link capacity, and to achieve the optimization objective by maximizing the system communication rate. Specifically, this invention solves the following problems:

[0080] 1. Determining the set of user devices for drone services: Determine which user devices each drone should provide services to in order to maximize the satisfaction of users' communication quality requirements.

[0081] 2. Determining the hovering position of each drone: Determine the optimal hovering position for each drone to maximize system communication speed and maintain stability.

[0082] 3. Determine the transmit power of each UAV: ​​Determine the transmit power of each UAV to achieve effective signal transmission and communication quality optimization.

[0083] 4. Determining the number of drones needed: Determine the required number of drones to meet communication needs and strike a balance between system efficiency and resource utilization.

[0084] To address the aforementioned problems, this invention proposes an algorithmic framework that decomposes the problem into two sub-problems and introduces substitution variables and mathematical transformations, transforming the problem into one that can be solved using continuous convex approximation methods. This algorithm can obtain near-optimal solutions while maintaining system stability. Compared to other methods, this method achieves higher system communication rates under the same conditions and can handle situations where other methods cannot yield feasible solutions.

[0085] The method of this invention can optimize a communication system with backhaul link capacity limitations based on multiple UAVs, improve communication quality and system performance, and has wide application and promotion value in the field of wireless communication.

[0086] This invention discloses a method for joint deployment and resource optimization based on multi-UAV-assisted wireless networks, such as... Figure 2 As shown, it includes the following steps:

[0087] Step 1: A drone-assisted wireless network was studied, in which a multi-antenna ground base station provides communication services in a disaster area through a group of drones. A time-division multiple access downlink communication link is used to combine multiple drones with user communication equipment. This wireless communication system consists of a ground base station, J drones, and K user communication equipment. User communication equipment communicates with the ground base station through the drones, while the drones act as airborne base stations, forwarding information from the ground base station via a backhaul link. In the design, this invention considers the communication quality requirements of the user communication equipment to ensure that their quality of service requirements are met, while also taking into account the capacity limitations of the wireless backhaul link between the base station and the drones.

[0088] Step 2: Under the constraints of backhaul link capacity limitations between the UAV and ground base stations, and user communication quality requirements,

[0089] A novel framework is proposed, which considers the deployment cost of UAVs and takes minimizing system cost as the optimization objective, and establishes an objective optimization problem.

[0090] Step 3: Decompose the optimization problem of minimizing system cost in Step 2 into two sub-optimization problems: the drone location deployment optimization problem and the drone power, communication bandwidth, user-drone association, and drone quantity optimization problem.

[0091] Step 4: Given the drone deployment locations, solve the optimization problem of minimizing system cost, optimizing drone power, communication bandwidth, user-drone association, and drone quantity.

[0092] Step 5: Given the drone power, communication bandwidth, user-drone association, and number of drones, solve the drone location deployment optimization problem within the optimization problem of minimizing system cost.

[0093] Step 6: Based on the solutions obtained in Steps 4 and 5, the cost minimization optimization problem of the multi-UAV assisted communication system is finally completed.

[0094] In this invention, the system comprises drones clustered within a specific area. There are J drones, labeled {1,2,…,J}; the set of user communication devices is... There are K devices in total, labeled {1, 2, ..., K}. Each of the K user devices has its own fixed ground location, and the three-dimensional coordinates of the k-th user device are given by: The goal is to deploy J drones to provide communication services to user equipment. The entire system employs time-division multiple access, considering downlink communication. The drones act as airborne base stations, forwarding information from ground base stations via backhaul links. The j-th drone hovers in a fixed position. It provides downlink communication for the set of user equipment it serves, and the coordinates of the ground base station are: The distance between the j-th drone and the k-th user equipment is denoted as d. j,k =||u j -u k ||2, i.e. u j and u k The Euclidean norm between them, similarly, the distance between the ground base station and the j-th UAV is denoted as d. 0,j =||u j -u0||2. The UAV and the ground base station are connected via a wireless backhaul link. The ground base station communicates with the UAV through a millimeter-wave channel, and the ground base station operates in the "massive multiple input multiple output (massive MIMO)" region. The beam gain at this time can be estimated as follows: N t N represents the number of antennas equipped in a ground base station. g Let J be the number of drones. Under the above conditions, the capacity of the wireless backhaul link between the drones and the ground base station can be expressed as:

[0095]

[0096] Among them, P GBS γ represents the transmit power of the ground base station, γ represents the environment-dependent backhaul link attenuation rate, and σ represents the transmit power of the ground base station. 2 Let g represent the noise power density of additive white Gaussian noise. Since the UAV is suspended in the air, its channel can be considered a line-of-sight air-to-ground channel, therefore the channel power gain g between the ground base station and the j-th UAV is... 0,j Specifically, it is expressed as follows:

[0097]

[0098] ρ0 is the channel power gain at the standard reference distance, and α is the path loss exponent. Similarly, the channel power gain between the j-th UAV and the k-th user equipment is denoted as:

[0099]

[0100] In this invention, a drone provides downlink communication to at least one user equipment using time-division multiple access, employing a binary variable x. j,k This indicates the allocation relationship between the drone and user equipment, a j Indicates the on / off status of the drone, α j,k This indicates the communication bandwidth allocation relationship between the drone and the user equipment. Specifically, a j =1 indicates that the j-th drone is in the active state. Conversely, if a = 1, then the j-th drone is active. j =0 indicates that the j-th drone is in a turned-off state. Specifically, x j,k =1 indicates that the k-th user equipment is assigned to the j-th drone for communication. Conversely, if x j,k = 0 indicates that the k-th user equipment does not belong to the service user equipment set of the j-th UAV. The achievable communication rate r from the j-th UAV to the k-th user equipment. k Represented as:

[0101]

[0102] p j This represents the transmit power of the j-th drone. Considering downlink communication among multiple drones, the user equipment will receive co-channel interference from other non-target drones. The combination represents the co-channel interference caused by other drones to the k-th user equipment. Multiple drones simultaneously conducting downlink communication require flexible adjustments to the communication connection between drones and user equipment, as well as the drone's transmit power, based on the actual distribution of user equipment to maximize the communication rate achievable by the user equipment. As a static airborne base station, the drone forwards information from the ground base station via a backhaul link. However, the wireless backhaul link has capacity limitations; the total communication rate from the j-th drone to its assigned set of user equipment cannot exceed the capacity of the wireless backhaul link between the j-th drone and the ground base station, mathematically expressed as:

[0103]

[0104] Each user device has its own communication quality requirements. Let the communication quality requirements of user equipment k be represented, then the following constraints apply:

[0105]

[0106] This indicates that the k-th user device needs to communicate with a certain drone, and the communication rate must be greater than [a certain value]. The optimization goal of the system is to determine the communication connection relationship between the UAV and the user equipment. j,k The hovering position of the drone. j And the transmission power p of the drone as an airborne base station j The optimization objective is:

[0107]

[0108]

[0109]

[0110]

[0111] C4:α jk ≤x jk,

[0112] C5:x jk ≤a j,

[0113]

[0114]

[0115] C8:a j ∈{0,1},

[0116] C9:x j,k ∈{0,1},

[0117]

[0118]

[0119]

[0120] Where, min indicates that the optimization objective is to minimize, st followed by expressions represents the constraints, and the subscripts j and k represent the j-th UAV and the k-th user equipment, respectively; r k Let k represent the communication rate of user equipment. The optimization objective is to maximize the sum of the communication rates of all user equipment. The optimization variables are: This represents the three-dimensional hovering position vector of the UAV, where the components are... These represent the coordinate values ​​on the X, Y, and Z axes in a three-dimensional coordinate system, respectively. Similarly, the constraints... and This indicates the lower and upper bounds of the coordinate vectors that can be selected, limiting the geographical range in which the drone can hover; C represents the communication quality requirements of user equipment k; j (d 0,j ) represents the upper limit of the backhaul link capacity between the UAV j and the ground base station, where d 0,j Let j be the physical distance between the drone and the ground base station; Represents the transmit power threshold of UAV j; ∑ j x j,k The constraint = 1 indicates that each user device can only connect to one drone. This problem is a mixed-integer nonlinear optimization problem, due to the binary variable x j,k This allows for the discretization of the feasible region of the constraints; the UAV's position u j Simultaneously with the backhaul link capacity C j (d 0,j ) and communication rate r k The interconnectedness of the constraints makes them non-convex, further hindering the determination of the optimal solution. To make the problem solvable, the block coordinate descent method is used to break it down into two smaller subproblems for solving.

[0121] Assuming the drone's location is fixed, the first optimization steps are to optimize the communication connection between the drone and the user equipment, the drone's power, the number of drones, and the drone's bandwidth allocation. The optimization problem then becomes...

[0122]

[0123]

[0124]

[0125]

[0126] C4:α jk ≤x jk,

[0127] C5:x jk ≤a j,

[0128]

[0129]

[0130] C8:a j ∈{0,1},

[0131] C9:x j,k ∈{0,1},

[0132] Since the left side (LHS) of constraint C1 and the right side (RHS) of constraint C2 are both non-convex functions, problem P1.1 remains non-convex. The transmission rate can be written as follows:

[0133]

[0134] By introducing the auxiliary variable m kj The above equation can be equivalently expressed as:

[0135]

[0136]

[0137] Due to the structure of the transmission rate, it can be directly converted into a function of p. j and m kj The difference between two concave functions. To introduce an auxiliary variable, this invention will... Defined as the lower bound of the second inequality above, we have:

[0138]

[0139] Constraint C1 can be expressed as

[0140] To solve constraint C2, this invention introduces an auxiliary variable z, and the formula can be equivalently expressed as follows:

[0141]

[0142]

[0143] Furthermore, the first inequality above can be equivalently rewritten as:

[0144] By introducing an additional variable m kj γ k , z, and x kj and a j By treating relaxation as a continuous variable, this invention can equivalently transform problem P1.1 into...

[0145]

[0146]

[0147]

[0148] C3-C7,

[0149] C8:0≤a j ≤1

[0150] C9: 0 ≤ xj,k ≤1

[0151]

[0152]

[0153]

[0154] use The above constraints can be equivalently expressed as:

[0155] 4(p j -m kj )+(x kj -p j ) 2 ≤(x kj +p j ) 2 ,

[0156]

[0157] Since problem P1.1.1 still has non-convex constraints, problem P1.1.1 remains non-convex. This invention uses a first-order approximation as their replacement function. The Taylor expansion approximation is as follows:

[0158]

[0159]

[0160]

[0161]

[0162] Where z (t) , This is a given value in the t-th iteration. (Function) It is a substitution function of the first-order Taylor expansion.

[0163] In the objective function, this invention introduces a penalty term to force the continuous variable x to... kj and a j Converges towards binary values ​​(0 or 1). The penalty period is defined as follows: The binary variable x... kj and a j Since relaxation is a continuous variable, this invention further introduces a penalty term. This ensures that when the penalty factors λ2 and λ3 are sufficiently large, the variable x... kj and a j It will converge to 0 or 1. Therefore, the final objective function becomes:

[0164]

[0165] The objective function is a convex function. By minimizing the upper bound of the objective function, we can maximize the original objective function. The upper bound of the objective function is expressed as:

[0166]

[0167] in The penalty term is at a given point and The first-order Taylor expansion. The specific form is as follows:

[0168]

[0169] Through the algebraic transformations described above, the optimization problem is transformed into a series of standard convex optimization problems. By using a continuous convex approximation algorithm, i.e., solving the standard convex optimization problem once using existing convex optimization problem solving tools, and then using the obtained solution as the initial solution in the next problem, iterating until convergence, the optimized communication connection between the UAV and the user equipment, the UAV power, the UAV bandwidth allocation, and the number of UAVs can be obtained.

[0170] Assuming the communication connection between the UAV and the user equipment, the UAV power, the UAV bandwidth allocation, and the number of UAVs are fixed, we can optimize the UAV hovering position using the communication connection, UAV power, UAV bandwidth allocation, and number of UAVs obtained above. The UAV position optimization problem can be expressed as:

[0171]

[0172]

[0173] P1.C2-P1.C3

[0174] P1.C10-P1.C12

[0175] The variable introduced is related to the objective function; at this point, the UAV's position u... j Exists in the communication rate expression r k and wireless backhaul link C j In the middle. By estimating r k and C j This non-convex problem can be transformed into a series of convex optimization problems that can be solved using continuous convex estimation. Specifically, r k It can be rewritten as:

[0176]

[0177] Introducing slack variable s kj ,v kj This reduces the distance between the drone and the user equipment. satisfy:

[0178]

[0179]

[0180] This allows for relaxation of r. j,k (u j ),Right now:

[0181]

[0182]

[0183] Based on this relaxation and first-order Taylor expansion, we can further give r j,k (u j The upper and lower bounds of r. j,k (u j The lower bound of ) is:

[0184]

[0185] It is the symbolic representation of that lower bound. and That is, r j,k (s j,k ,v j,k The Taylor expansion point of ). By r j,k (u j Substitute the lower bound of ) into the constraint These two non-convex constraints can be transformed into standard convex constraints. To handle the non-convexity of constraint P1:C2, the backhaul link capacity C is further processed. j (u j ), and give C again j (u j The expression for ) is:

[0186]

[0187] Since terrestrial base stations operate in massive multi-antenna mode, the signal-to-noise ratio can be considered... Therefore, it can be estimated as Further expand the backhaul link as follows:

[0188]

[0189] in and This is a term independent of the variable, the drone's position. We introduce the variable t. j , so that:

[0190]

[0191] Then there is The present invention can then provide... and d 0,j The first-order Taylor expansions are denoted as follows: and The specific form is as follows:

[0192]

[0193]

[0194] in, This refers to the expansion point of the Taylor expansion, and also the UAV's position coordinates at the t-th iteration in the continuous convex estimation. Through the variables and mathematical transformations introduced above, the UAV position optimization problem can be expressed as follows:

[0195]

[0196]

[0197]

[0198]

[0199]

[0200]

[0201] P1.C10-P1.C12

[0202] The above problem is a standard convex optimization problem. By solving for the drone's position in the t-th iteration, the obtained drone position is... Using this as the Taylor expansion point for the next iteration until convergence, the optimized drone position can be obtained.

[0203] The iterative solution of the convex optimization problem obtained by the continuous convex approximation algorithm continues until the communication rate of the user equipment converges. That is, it can solve for all the variables to be solved. For example... Figure 3 As shown, by iteratively solving the above two sub-problems until the objective function converges, the cost minimization optimization of the communication system based on multi-UAV assisted communication is completed.

[0204] Example:

[0205] As shown in Table 1, the parameter settings of the above system are given. Parameters without specific values ​​will be used as comparison parameters. If there are multiple different values, detailed explanations will be provided separately.

[0206] Table 1 Parameter Setting Table

[0207]

[0208]

[0209] This problem is an NP-hard problem in optimization. For the proposed NP-hard problem, it can be solved using block coordinate descent (a type of optimization algorithm that expresses the objective function as a function of multiple variables, optimizing only one variable until the optimal solution is reached) and successful convex approximation (which iteratively solves a series of convex optimization problems similar to the original problem; when the final convergence condition is met, the obtained solution can be approximated as the solution to the original problem).

[0210] First, to verify the performance of the proposed algorithm in minimizing the total system cost, this invention considers a downlink transmission scenario where multiple drones use wireless backhaul, and the base station is located at [0,0,30]. Figure 4 This invention demonstrates the block coordinate descent iterative process of the proposed algorithm under different numbers of users. The invention deployed 20, 30, 40, 50, and 60 users respectively within a 500×500 meter square area. Assuming the static power consumption λ of each UAV is 1, the invention observes that the proposed algorithm converges rapidly within several iterations.

[0211] Figure 5 illustrates the deployment of drones using an algorithm, taking into account a specific user location distribution. Arrows indicate the drone's initial and final locations. Circles without arrows represent inactive drones. Interrelated drones are connected to users by dashed lines.

[0212] Figure 5a and Figure 5bTwo-dimensional and three-dimensional representations of user locations and drone deployment diagrams are provided. In this scenario, 45 users, indicated by crosses, are randomly distributed within a 500×500 meter square area. To initialize a virtual drone set, this invention employs a grid discretization method, selecting seven grid intersections as initial positions. The initial positions of these drones are represented by hollow circles, while the optimal positions are represented by solid circles. By executing the algorithm, this invention can determine the actual set of drones that need to be deployed, their correspondence with users, and the optimal deployment positions. After the algorithm is executed, drones that do not need to be deployed are represented by black hollow circles. Notably, these drones are far from the drones that need to be deployed. Conversely, the solid circles corresponding to the crosses represent drones that need to be deployed, serving all users indicated by the crosses. As can be clearly seen from Figure 5, the deployed drones (solid circles) and the users they serve (crosses) are accurately represented. Conversely, the black hollow circles represent drones that do not need to be deployed, highlighting the effectiveness of the algorithm in determining the optimal drone deployment strategy.

[0213] In contrast, this invention uses the MeanShift drone deployment algorithm. The MeanShift method is a clustering method that cannot adjust drone deployment based on user communication quality (QoS) requirements and backhaul capacity. A comparison of MeanShift clustering methods shows the number and allocation of drones deployed under the same parameters and user locations, such as... Figure 6 As shown, it can be observed that in the MeanShift clustering method, some user locations are relatively isolated, resulting in one drone serving two users, which wastes communication resources.

[0214] To further evaluate the performance of the proposed method, this invention underwent 100 iterations using a randomly generated user distribution and calculated the average values ​​of relevant metrics. This evaluation allows for a comprehensive analysis of the effectiveness of the proposed method. Based on the algorithm and the MeanShift clustering method used to determine the optimal drone deployment cost for different numbers of users, this invention presents the results of drone deployment cost and power consumption in the form of a bar chart, as shown below. Figure 7 As shown, with λ=1, the diagram illustrates the deployment cost and power consumption of drones using different deployment methods (the curve represents power consumption, and the bar chart represents the number of drones). The results of this invention demonstrate that, compared to the MeanShift clustering method, the algorithm of this invention consistently achieves lower drone deployment costs. Furthermore, the drone power consumption is dynamically adjusted based on the number of users, thereby reducing the overall system cost compared to the MeanShift method. These results confirm the effectiveness of the proposed method in achieving resource optimization while meeting user QoS requirements and backhaul link capacity limitations.

[0215] In its specific implementation, this application provides a computer storage medium and a corresponding data processing unit. The computer storage medium is capable of storing a computer program, which, when executed by the data processing unit, can run the invention's content regarding the joint deployment and resource optimization method based on a multi-UAV assisted wireless network, as well as some or all of the steps in various embodiments. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0216] Those skilled in the art will clearly understand that the technical solutions in the embodiments of the present invention can be implemented using computer programs and their corresponding general-purpose hardware platforms. Based on this understanding, the technical solutions in the embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of computer programs, i.e., software products. These computer program software products can be stored in a storage medium and include several instructions to cause a device containing a data processing unit (which may be a personal computer, server, microcontroller, MUU, or network device, etc.) to execute the methods described in various embodiments or certain parts of the embodiments of the present invention.

[0217] This invention provides a concept and method for the joint deployment and resource optimization of a multi-UAV-assisted wireless network. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment of the invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A method for joint deployment and resource optimization based on multi-UAV assisted wireless networks, characterized in that, Includes the following steps: Step 1: Establish a communication system model based on a multi-UAV assisted wireless network; Step 2: For the communication system model, design optimization objectives and establish optimization problems; Step 3: Decompose the problem of minimizing system cost in Step 2 into two sub-optimization problems: the drone location deployment optimization problem and the drone setup optimization problem. Step 4: Given the deployment locations of the drones, solve the drone setup optimization problem; Step 5: Given the drone settings, solve the drone location deployment optimization problem; Step 6: Based on the solutions obtained in Steps 4 and 5, complete the problem of minimizing system cost, that is, realize the joint deployment and resource optimization based on multi-UAV assisted wireless network; The communication system model described in step 1 is as follows: A multi-antenna ground base station provides communication services within a region via a group of drones; it employs a time-division multiple access downlink communication link to combine the drones with user communication equipment; the communication system consists of a ground base station, A drone and It consists of individual user communication devices, specifically including: The drone collection in the communication system is There are a total of Each drone is labeled as follows: The user communication equipment set is There are a total of Each device is labeled as follows: drones With user equipment If a communication connection needs to be established, it is marked as such. Otherwise, mark as ; Indicates the first The operational status of each drone, whether it is turned on or off. Indicates the first The drone and the first Communication bandwidth of each user device; The optimization objective described in step 2 is to minimize system cost, and the communication connection between the UAV and user equipment is determined based on the optimization objective. Communication bandwidth allocation between UAVs and user equipment 3D hovering position of the drone Transmission power of drones as airborne base stations ; The problem of minimizing system cost is defined as an optimization model P1, as follows: ； ； ； ； ； ； ； ； ； ； ； ； ； in, This indicates that the optimization objective is to minimize, The following expression indicates a constraint condition, indicated by the subscript. and They represent the first The drone and the first One user device; Indicates user equipment The communication rate is optimized with the goal of minimizing system cost. In constraint C1 Indicates user equipment Communication quality requirements; In constraint C2 Indicates drone The upper limit of backhaul link capacity with ground base stations, of which For drones Physical distance from ground base stations; In constraint C3 Indicates drone The transmit power threshold; Constraint C4 means that when the first The user and the first When a drone is not connected, the allocated bandwidth score =0; Constraint C5 indicates that the drone providing the service is active; Constraint C6 means that each user is served by only one drone; Constraint C7 states that only active drones can connect to the user, and the communication bandwidth of each drone is limited. The sum is less than 1; Constraints C8 and C9 represent and It is binary; Optimization variables in constraints C10-C12 This represents the three-dimensional hovering position vector of the UAV, where the components are... These represent the coordinate values ​​on the X, Y, and Z axes in a three-dimensional coordinate system; in the constraints and This represents the lower and upper bounds of the coordinate vector, used to limit the geographical range of where the drone can hover.

2. The method for joint deployment and resource optimization based on multi-UAV assisted wireless networks according to claim 1, characterized in that, The optimization objective described in step 2 is to minimize the system cost, and the optimization problem is the system cost minimization problem.

3. The method for joint deployment and resource optimization based on multi-UAV assisted wireless networks according to claim 2, characterized in that, The drone settings described in step 3 include: drone power, communication bandwidth, user association with drone, and number of drones.

4. The method for joint deployment and resource optimization based on multi-UAV assisted wireless networks according to claim 3, characterized in that, The drone setup optimization problem described in step 3, namely, given the drone's location, yields sub-problems P1.1 for optimizing drone power, communication bandwidth, user-drone association, and the number of drones, as detailed below: ； ； The constraints of subproblem P1.1 are constraints C1-C9 in optimization model P1, where the position of the UAV is known.

5. The method for joint deployment and resource optimization based on multi-UAV assisted wireless networks according to claim 4, characterized in that, The drone location deployment optimization problem described in step 3, namely the sub-problem P1.2 of optimizing drone location given drone power, communication bandwidth, user-drone association, and the number of drones, is as follows: ； ； ； ； The constraints of subproblem P1.2 are constraints C2-C3 and C10-C12 in optimization model P1, where constraint C1... Indicates user equipment The lower bound of the rate is given, and the drone power, communication bandwidth, user-drone association, and number of drones are known.

6. The method for joint deployment and resource optimization based on multi-UAV assisted wireless networks according to claim 5, characterized in that, The specific method for solving the UAV setup optimization subproblem in step 4 includes: For communication connection variables and drone switch state variables Introduce a penalty term into the objective function: as well as This causes the above optimization variables to converge to A set of binary variables, in the penalty term This is the penalty coefficient; For other non-convex constraints in subproblem P1.1, a continuous convex approximation algorithm is adopted, using a first-order Taylor expansion as a substitute function for the non-convex function to transform the non-convex constraints into convex constraints. This formulation is further transformed into iteratively solving a series of standard convex optimization problems to obtain the solution to subproblem P1.

1.

7. The method for joint deployment and resource optimization based on multi-UAV assisted wireless networks according to claim 6, characterized in that, The specific methods for solving the UAV location deployment optimization problem in step 5 include: For a series of non-convex constraints in subproblem P1.2, a continuous convex approximation algorithm is adopted, using a first-order Taylor expansion as a substitute function for the non-convex function to transform the non-convex constraints into convex constraints. This formulation is further transformed into iteratively solving a series of standard convex optimization problems, thereby obtaining the solution to subproblem P1.

2.

8. The method for joint deployment and resource optimization based on multi-UAV assisted wireless networks according to claim 7, characterized in that, Step 6 involves solving the problem of minimizing system cost by using the block coordinate descent method to perform alternating iterative optimization on the results obtained in steps 4 and 5 to obtain the minimum system cost.

Images