Position deployment and resource allocation method for multi-unmanned aerial vehicle communication

Abstract

The invention discloses a position deployment and resource allocation method for multi-unmanned aerial vehicle communication. The method comprises the following operation steps: S1, initializing parameters and a model; s2, modeling: modeling for unmanned aerial vehicle position deployment and resource allocation problems; s3, splitting the problem by using a block coordinate descent method, and splitting the mixed integer non-convex problem by using the block coordinate descent method; s4, user association and the three-dimensional position of the unmanned aerial vehicle, wherein the user association and the three-dimensional position of the unmanned aerial vehicle are obtained through an ant lion algorithm; s5, transmitting power of the unmanned aerial vehicle, and optimizing the transmitting power of the unmanned aerial vehicle by using a successive convex approximation technology; and S6, an iterative optimization algorithm is carried out. Under the condition of multi-unmanned aerial vehicle deployment, the transmission rate requirement of the ground users can be met at the same time, the position deployment and resource allocation of the multiple unmanned aerial vehicles are optimized, the transmission rate is maximized, the system throughput can be improved, and the lowest rate requirement of all the ground users can be met.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle communication, and in particular to a location deployment and resource allocation method for multi-unmanned aerial vehicle communication. Background Art

[0002] Drones are unmanned aircraft that are controlled by radio remote control equipment and self-contained program control devices. With the rapid development of wireless communication technology, future communication networks need to meet users' complex and changing communication requirements. General wireless communication systems are mainly composed of base stations fixed on the ground, but complex communication environments will limit the development of communication networks. Drones are easy to deploy and flexible, which can effectively solve this problem. Drones can act as air base stations or relays to help ground users transmit information. As air base stations, drones can expand the coverage of wireless networks and improve wireless communication services for ground users.

[0003] In most scenarios, drones are considered to be deployed as low-altitude platforms. In drone communication systems, the communication channels between drones and between drones and ground users are different from traditional communication channels, and the channel characteristics are more complex. The channel model of drone communication is different from that of traditional ground communication in the following ways. First, due to the high mobility of drones, the air-to-ground and air-to-air communication channel models are non-fixed and time-varying, and the Doppler shift caused by the movement of drones needs to be considered. Secondly, the deployment height of drones is relatively high, and the communication between drones and ground terminal users is not blocked, which increases the probability of line-of-sight transmission links. Drones are usually deployed as low-altitude platforms, and there are tall buildings and trees around them, as well as the influence of the fuselage caused by the structural design and rotation of the drone itself. The communication channel between drones and ground users will have a component of non-line-of-sight link transmission.

[0004] A journal article titled "Joint Optimization Method of User Matching and Spectrum Resources in Multi-UAV Assisted Communications" mentions a drone location deployment scheme based on the K-means clustering algorithm. The basic principle of drone location deployment based on the K-means clustering algorithm is: first, input the number of drones and the location coordinates of ground users, randomly generate M different seeds from the user locations as cluster centers, and repeat this operation; traverse all ground users, find the cluster center with the smallest distance to the ground users based on the Euclidean distance, and assign the users to the cluster; then, calculate the class center of each cluster, the horizontal coordinate of the class center is the average of the horizontal coordinates of all elements in the cluster, and the vertical coordinates are similar, and move the seed to its class center to obtain a new coordinate seed; finally, if the deviation between the newly obtained seed coordinates and the previous seed coordinates is less than a certain threshold, the process ends; otherwise, the loop returns.

[0005] However, according to the above-mentioned "Joint Optimization Method of User Matching and Spectrum Resources in Multi-UAV Assisted Communications", for the UAV-user association sub-problem, the K-Means Algorithm (KMA) is used for optimization to ensure that the distance between each ground user and its service UAV is minimized, minimizing the path loss of the transmission rate. The problem is that K-means is very sensitive to the selection of the initial cluster center, which may lead to local optimality. Although there are improved algorithms (such as K-means++) that can alleviate this problem, it still cannot be completely solved. This algorithm is suitable for simple scenarios. In multi-UAV deployment, due to the presence of various interferences, the actual situation is more complicated, which affects the clustering results. K-means needs to specify the number of clusters in advance. Improper selection of initial values ​​and cluster centers may lead to unsatisfactory clustering results.

[0006] Therefore, in the existing technology, the conditions such as the deployment location of drones, resource allocation and communication channels have a great impact on drone communications, and it is difficult to simultaneously meet the transmission rate requirements of ground users. Summary of the invention

[0007] The technical problem to be solved by the present invention is to overcome the defect of low transmission rate of multi-UAV communication in the prior art and to provide a location deployment and resource allocation method for multi-UAV communication.

[0008] The present invention solves the above technical problems through the following technical solutions:

[0009] The present invention provides a location deployment and resource allocation method for multi-UAV communication, the method comprising the following steps:

[0010] S1, parameter and model initialization;

[0011] S2, modeling, modeling the UAV location deployment and resource allocation problems;

[0012] S3, block coordinate descent method to split the problem, using block coordinate descent method to split the mixed integer non-convex problem;

[0013] S4, user association and drone 3D position, the ant lion algorithm obtains the drone user association and 3D position;

[0014] S5, UAV transmission power, using successive convex approximation technology to optimize the UAV transmission power;

[0015] S6, iterative optimization algorithm to obtain a suboptimal solution.

[0016] In this technical solution, a downlink communication system for multi-UAV-assisted ground user communications is constructed. User matching, UAV position and transmission power are jointly optimized under the condition of channel interference, achieving the goal of maximizing the minimum user rate.

[0017] Preferably, in step S1, the drone serves as a temporary base station to serve all ground users within the coverage area. One drone can cover multiple ground users, and the drone establishes communication with ground users via a downlink.

[0018] Preferably, the models established in step S2 are a channel model, a communication model and an optimization model.

[0019] Preferably, the air-to-ground link in the channel model adopts a hybrid channel model, and the hybrid channel model is a line-of-sight link and a non-line-of-sight link.

[0020] Preferably, in the step S3, the mixed integer non-convex problem is split into two sub-problems, and the two sub-problems are respectively the user association and drone two-dimensional deployment problem and the drone transmission power problem.

[0021] Preferably, the specific operation method of step S4 is:

[0022] Step 1: User association and two-dimensional deployment of drones;

[0023] Step 2: Determine the altitude of the drone;

[0024] Step 3: Jointly optimize user association and drone location deployment.

[0025] Preferably, the improved ant lion algorithm is used for optimization in step three, and the specific operation process is as follows:

[0026] A1. Start;

[0027] A2. Initialize various parameters;

[0028] A3. Randomly generate population positions, calculate antlion fitness, and select elite antlions;

[0029] A4. Select antlions through a roulette mechanism;

[0030] A5, ants randomly walk around the ant lion;

[0031] A6. Update the ant position;

[0032] A7. Calculate the objective function value and select the antlion with the highest fitness as the elite antlion;

[0033] A8, determine whether the number of iterations reaches the upper limit, if yes, execute step A9, if no, execute step A4;

[0034] A9. End.

[0035] Preferably, in step S5, SCA technology is used to optimize the transmission power of the UAV.

[0036] Preferably, the specific operation steps of the iterative optimization algorithm in step S6 are:

[0037] B1, parameter initialization, the number of iterations of the drone’s position and resource allocation initialization;

[0038] B2, the cycle starts;

[0039] B3. Given, use Algorithm 1 to solve the user association and drone location problems and obtain the optimal solution for the iteration;

[0040] B4. Given {A (t+1) , V (t+1)}, use CVX to solve the UAV transmission power problem and get the optimal solution P for the t+1th iteration (t+1) ;

[0041] B5, t = t + 1, calculate {A (t+1) , P (t+1) , V (t+1)}, until the relative increase of the minimum value of all user rates is lower than the threshold λ or t>T;

[0042] B6. Get the user association A, drone position V and transmission power P.

[0043] Preferably, in step S5, a first-order Taylor expansion and an approximate method are used to transform a non-convex problem into a convex problem, and a standard convex optimization tool is used to solve the convex problem.

[0044] On the basis of being in accordance with the common sense in the art, the above-mentioned preferred conditions can be arbitrarily combined to obtain the preferred embodiments of the present invention.

[0045] The positive and progressive effects of the present invention are:

[0046] The present invention constructs a downlink communication system for multi-UAV-assisted ground user communication, jointly optimizes user matching, UAV position and transmission power under the condition of channel interference, and achieves the goal of maximizing the minimum user rate;

[0047] In order to solve the optimization problem with non-convex objective function and constraints, a two-stage alternating iterative optimization algorithm is proposed, which can obtain the suboptimal solution of the optimization problem; specifically, the block coordinate descent method is used to convert the optimization problem into two sub-problems; first, the user association and drone location deployment problems are solved by the ant lion algorithm (ALO); in addition, the height of the drone is calculated by the formula. It can achieve full coverage of ground users and accelerate the convergence speed of the algorithm;

[0048] Since the power optimization problem of the UAV is non-convex, the first-order Taylor expansion and approximation method are used to transform the non-convex problem into a convex problem, and the convex problem is solved using standard convex optimization tools; ultimately, the approximate optimal solution to the optimization problem is obtained.

[0049] The present invention constructs an optimization problem of maximizing user rate under the constraints of user association, UAV location deployment and transmission power, while taking channel interference into consideration. It can significantly improve the minimum transmission rate of ground users in multi-UAV communications, achieve full coverage of all ground users by multiple UAVs, and optimize the transmission power of UAVs. The proposed two-stage optimization algorithm can effectively improve the transmission rate of all users, verifying the effectiveness of the proposed method.

[0050] The present invention can simultaneously meet the transmission rate requirements of ground users under the condition of multiple UAV deployment, optimize the position deployment and resource allocation of multiple UAVs, maximize the transmission rate, improve the system throughput, and meet the minimum rate requirements of all ground users. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 The present invention is a schematic diagram of the overall process of the location deployment and resource allocation method for multi-UAV communication according to an embodiment of the present invention.

[0052] Figure 2 for Figure 1 The flowchart of the specific operation method of step S4 of the location deployment and resource allocation method for multi-UAV communication is shown.

[0053] Figure 3 for Figure 1 The flowchart of the specific operation method of step three of the location deployment and resource allocation method for multi-UAV communication is shown.

[0054] Figure 4 for Figure 1 The flowchart of the specific operation method of step S6 of the location deployment and resource allocation method for multi-UAV communication is shown.

[0055] Figure 5 for Figure 1 Schematic diagram of the system model of the location deployment and resource allocation method for multi-UAV communication shown.

[0056] Figure 6 for Figure 1 Schematic diagram of the coverage area of ​​drones showing the location deployment and resource allocation method for multi-drone communication.

[0057] Figure 7 for Figure 1 Schematic diagram of ground user distribution of location deployment and resource allocation method for multi-UAV communication shown.

[0058] Figure 8 for Figure 1 Schematic diagram of multi-UAV deployment of location deployment and resource allocation method for multi-UAV communication Figure 1 .

[0059] Fig. 9 for Figure 1 Schematic diagram of multi-UAV deployment of location deployment and resource allocation method for multi-UAV communication Figure 2 .

[0060] Fig.10 for Figure 1 The iterative schematic diagram of the ant lion algorithm for location deployment and resource allocation method for multi-UAV communication is shown.

[0061] Fig.11 for Figure 1 The iterative schematic diagram of the overall algorithm of the location deployment and resource allocation method for multi-UAV communication is shown.

[0062] Fig.12 for Figure 1 Schematic diagram of the variation of user minimum rate with the number of drones for the location deployment and resource allocation method for multi-UAV communication shown.

[0063] Fig.13 for Figure 1 Schematic diagram of the variation of the minimum user rate with the number of users for the location deployment and resource allocation method for multi-UAV communication shown.

[0064] Fig.14 for Figure 1 The diagram shows how the minimum user rate of the location deployment and resource allocation method for multi-UAV communication changes with the number of users and UAVs. DETAILED DESCRIPTION

[0065] The present invention is further described below by way of examples, but the present invention is not limited to the scope of the examples.

[0066] Figures 1 to 14The figure shows a schematic diagram of an embodiment of a method for position deployment and resource allocation for multi-UAV communication of the present invention. The method for position deployment and resource allocation for multi-UAV communication comprises the following steps:

[0067] S1, parameter and model initialization;

[0068] S2, modeling, modeling the UAV location deployment and resource allocation problems;

[0069] S3, block coordinate descent method to split the problem, using block coordinate descent method to split the mixed integer non-convex problem;

[0070] S4, user association and drone 3D position, the ant lion algorithm obtains the drone user association and 3D position;

[0071] S5, UAV transmission power, using successive convex approximation technology to optimize the UAV transmission power;

[0072] S6. Iterate the optimization algorithm to obtain a suboptimal solution.

[0073] In this technical solution, a downlink communication system for multi-UAV-assisted ground user communications is constructed. User matching, UAV position and transmission power are jointly optimized under the condition of channel interference, achieving the goal of maximizing the minimum user rate.

[0074] Ant lion optimizer (ALO) is an optimization algorithm based on the behavior of ant lions digging funnel-shaped traps to prey on ants. It is a meta-heuristic algorithm. Ant lions dig a conical pit in the sand and hide at the bottom of the pit waiting for prey to fall in. Once the ant lion realizes that the prey is in the trap, it will use its huge jaws to throw sand to the edge of the pit to force the prey to slide into the bottom of the pit and then prey.

[0075] Successive convex approximation (SCA) technology is continuous convex approximation. Its core idea is to approximate the original non-convex optimization problem or the problem that is difficult to solve. By iteratively solving a series of convex problems similar to the original problem, an approximate solution to the original optimization problem is obtained. This algorithm is widely used in solving UAV trajectory planning problems. This technology is usually used to convert non-convex problems into convex problems.

[0076] In the step S1, the drone acts as a temporary base station and serves all ground users within the coverage area. One drone can cover multiple ground users, and the drone establishes communication with ground users through a downlink.

[0077] The scenario of step S1 is as follows Figure 5As shown in the figure, consider the drone as a temporary base station, serving all ground users within the coverage area. One drone can cover multiple ground users, and the drone establishes communication with the ground users through the downlink. There are M drones in the air area and K users on the ground. They are represented as the set M = {1,2,...,M} and the set K = {1,2,...,K} respectively. The three-dimensional Cartesian coordinates of the kth ground user and the mth drone are w k ={x k ,y k ,h k} and v m ={x m ,y m ,h m}. Assume that ground users are distributed in h k = 0 horizontal plane, where h min <h m <h max , where h min and h max denote the minimum and maximum allowed altitudes of the mth UAV, respectively. Without loss of generality, q k ={x k ,y k} is the horizontal coordinate of the kth ground user, u m ={x m ,y m} is the horizontal coordinate of the mth UAV, and the downlink transmission power of the mth UAV satisfies p min <p m <p max , p min and p max Indicates the minimum and maximum transmission power. Due to the use of time division multiple access, users in the drone occupy a sub-channel exclusively, and there is no interference between users. Multiple drones share a channel, so there is interference between drones.

[0078] The models established in step S2 are a channel model, a communication model and an optimization model.

[0079] The air-to-ground link in the channel model adopts a hybrid channel model, and the hybrid channel model is a line-of-sight link and a non-line-of-sight link.

[0080] The air-to-ground link in the channel model uses a hybrid channel model, namely line-of-sight (LoS) and non-line-of-sight (NLoS) links. The path loss can be expressed as formula 1 and formula 2 respectively.

[0081] Formula 1:

[0082] Formula 2:

[0083] Where: η LoS and η NLoS They represent the additional losses caused by LoS and NLoS links respectively; f c is the carrier frequency; c is the speed of light; d k,m represents the distance between the mth UAV and the kth ground user, as shown in Formula 3.

[0084] Formula 3:

[0085] The present invention adopts a probabilistic link loss model, and the line-of-sight link probability between the UAV m and the ground user k is shown in Formula 4.

[0086] Formula 4:

[0087] Where: a, b are constants related to environmental conditions; h m is the altitude of the drone; r k,m is the horizontal distance between the mth UAV and the kth ground user, as shown in Formula 5.

[0088] Formula 5:

[0089] The probability of a non-line-of-sight link can then be expressed as Formula 6.

[0090] Formula 6:

[0091] Therefore, the average path loss between UAV m and ground user k is shown in Equation 7.

[0092] Formula 7:

[0093] A user association matrix Α is defined in the communication model. Each element α in the matrix k,m ∈{0,1} represents the association between UAV m and ground user k. If α k,m =1, indicating that ground user k is provided with communication services by UAV m, otherwise α k,m = 0. In addition, each user is provided with communication services by at most one drone, and constraints are generated to meet the conditions, as shown in formulas 8 and 9.

[0094] Formula 8:

[0095] Formula 9:

[0096] The channel gain between UAV m and ground user k is shown in formula 10.

[0097] Formula 10:

[0098] Therefore, the corresponding received signal to interference plus noise ratio at ground user k can be expressed as formula 11.

[0099] Formula 11:

[0100] Where: 2 is the power of the additive white Gaussian noise at the receiver; the first term in the denominator represents the co-channel interference caused by the transmissions of other drones. m represents the transmission power of the mth UAV, p min <p m <p max , p min and p max represents the minimum and maximum transmit power. Therefore, the achievable rate R for ground user k served by UAV m is k,m , as shown in Formula 12.

[0101] Formula 12:

[0102] Therefore, the rate of ground user k is shown in formula 13.

[0103] Formula 13:

[0104] In the optimization model, in order to maximize the minimum rate of all ground users, an optimization problem is proposed. The drone allocates communication services to all ground users and achieves coverage for all users. Based on this deployment condition, The position matrix corresponding to the three-dimensional Cartesian coordinates of the drone, Corresponding to the association matrix associated with the drone, is the power allocation matrix. In order to facilitate the solution, this paper introduces represents the minimum rate of all ground users. Therefore, the optimization problem can be expressed as maximizing η . As shown in Formula 14.

[0105] Formula 14:

[0106]

[0107] In the above formula, C1 means that at least the minimum transmission rate of ground user k is met. C2 and C3 represent the user association constraints of UAV communication, ensuring that each user is served by one UAV. C4 means that the horizontal distance between the UAV and the user does not exceed the coverage radius of the UAV. C5 is the coverage radius constraint of the UAV. C6 is the transmission power constraint of the UAV.

[0108] Due to the non-convex constraints in the C1 constraint and the integer constraints in the C3 constraint, the optimization problem is a mixed integer non-convex problem. In order to facilitate the solution, the problem needs to be converted into sub-problems for solution.

[0109] In the S3 step, the mixed integer non-convex problem is split into two sub-problems, which are the user association and UAV two-dimensional deployment problem and the UAV transmission power problem.

[0110] Due to the non-convex constraints and integer constraints in the constraints, the optimization problem is a mixed integer non-convex problem. In order to facilitate the solution, the problem needs to be converted into two sub-problems to solve.

[0111] In order to solve the above non-convex optimization problem, an iterative algorithm based on block coordinate descent (BCD) is proposed to obtain a suboptimal solution to the original problem. The original problem is decomposed into two sub-problems. The user association and drone two-dimensional deployment problem is to optimize the user association and drone position under the given drone transmission power. The drone transmission power problem is to optimize the drone transmission power under the given user association and drone position. In addition, the drone transmission power problem is non-convex and can be converted into a convex problem using the continuous convex approximation method, and then solved using the standard convex optimization tool CVX.

[0112] The specific operation method of the S4 step is:

[0113] Step 1: User association and two-dimensional deployment of drones;

[0114] Step 2: Determine the altitude of the drone;

[0115] Step 3: Jointly optimize user association and drone location deployment.

[0116] In step 1, a two-dimensional deployment problem is modeled by placing multiple circles of different sizes. Each user is served by a UAV. Assuming that the coverage radius of the UAV is R, it is equivalent to finding a suitable horizontal position for each UAV and which user the UAV provides communication service for, that is, u m and A, to cover as many users as possible. An improved ant lion algorithm is proposed to solve the user association and drone location deployment problems. First, the ant lion algorithm is introduced.

[0117] Random walk is used to simulate the movement range of ants, as shown in Formula 15.

[0118] X(t)=[0, cumsum(2r(t1)-1), cumsum(2r(t2)-1),

[0119] Formula 15:···,cumsum(2r(t n )-1)]

[0120] Where: cumsum represents the calculation of cumulative sum; t is the walking step length; n is the maximum number of iterations. r(t) is defined as Formula 16.

[0121] Formula 16:

[0122] rand is a random number uniformly distributed in the range [0,1].

[0123] In order to ensure that the ants randomly walk within the search range, the following equation is used for normalization. As shown in Formula 17.

[0124] Formula 17:

[0125] Where: b i and a i They represent the upper and lower limits of the random walk in the ith variable respectively; and They represent the upper and lower limits of the i-th variable at the t-th iteration respectively.

[0126] The traps built by ant lions affect the random walk of ants in the search range, which can be expressed as Formula 18.

[0127] Formula 18:

[0128] Where: d t and c t Respectively represent the upper and lower limits of the variable at the tth iteration; represents the position of the jth antlion at the tth iteration. The ant’s wandering boundary is shown in Formula 19.

[0129] Formula 19:

[0130] Where I is defined as Formula 20.

[0131] Formula 20:

[0132] In the figure: t is the current iteration number; T is the maximum iteration number; v increases in a step-by-step manner as the number of iterations increases.

[0133] If the objective function value of the ant is better than that of the selected antlion, it can be expressed as Formula 21.

[0134] Formula 21:

[0135] The formula for updating the ant's position is shown in Formula 22.

[0136] Formula 22:

[0137] Where: is the position of the antlion chosen by the ants around the roulette wheel at the tth iteration, is the position of the elite ant lion after random walk at the tth iteration.

[0138] In the step of determining the height of the drone, the coverage area of ​​the drone is as follows: Figure 6 As shown. The horizontal position and coverage radius of the drone have been determined, and the next goal is to find the optimal altitude of the drone. From the formula of path loss, it can be observed that the path loss between the drone and the user is a function of the horizontal position and altitude. For a given path loss, the elevation angle is defined as By solving There exists an elevation angle θ that maximizes the radius max , the elevation angle of the drone satisfies Formula 23. The altitude of the drone is solved as shown in Formula 24.

[0139] Formula 23:

[0140] Formula 24: h m =Rtan(θ max )

[0141] In order to solve the problem of user association and drone location deployment, an improved ant lion algorithm is used for optimization. First, the ant lion position is randomly generated by initialization, and the position of the ground user is randomly distributed in the area. The fitness value of the ant lion is calculated, and the fitness value of the ant lion is sorted to obtain the position of the elite ant lion. Secondly, in the random walk strategy, an ant lion is randomly selected according to the roulette strategy, and the position of the ant lion is calculated after walking around the roulette and the elite ant lion. The ant is checked for the boundary and the fitness value of the ant is calculated. The position of the elite ant lion is updated and recorded as the global optimal value to obtain the optimal user association and the horizontal position of the drone. Finally, the height of the drone is calculated according to the formula.

[0142] In step 3, the improved ant lion algorithm is used for optimization, and the specific operation process is as follows:

[0143] A1. Start;

[0144] A2. Initialize various parameters;

[0145] A3. Randomly generate population positions, calculate antlion fitness, and select elite antlions;

[0146] A4. Select antlions through a roulette mechanism;

[0147] A5, ants randomly walk around the ant lion;

[0148] A6. Update the ant position;

[0149] A7. Calculate the objective function value and select the antlion with the highest fitness as the elite antlion;

[0150] A8, determine whether the number of iterations reaches the upper limit, if yes, execute step A9, if no, execute step A4;

[0151] A9. End.

[0152] In the step S5, the SCA technology is used to optimize the transmission power of the UAV.

[0153] Given the optimal solution of user association and drone position obtained in the first stage, the optimization sub-problem of drone transmission power is as shown in Formula 25.

[0154] Formula 25:

[0155]

[0156] Since the constraint C1 is non-convex, this part is solved based on the SCA technique. In order to use the successive convex approximation technique to convert the non-convex constraint into a convex constraint, first write the formula related to the variable P in the form of two functions subtracted, as shown in Formula 26.

[0157] Formula 26:

[0158] The second term in the formula is shown in Formula 27.

[0159] Formula 27:

[0160] Obviously, Formula 26 is non-convex, and the first-order Taylor expansion of any concave function has a global upper limit at any point. Therefore, in the tth iteration, the UAV transmission power is defined, and the corresponding convex upper bound at the local point is as shown in Formula 28.

[0161] Formula 28:

[0162] In the formula,

[0163] Formula 29:

[0164] Therefore, the UAV transmission power problem can be transformed into Formula 30.

[0165] Formula 30:

[0166] The optimization problems and constraints mentioned above are all convex and can be solved using convex optimization tools (CVX).

[0167] The specific operation steps of the iterative optimization algorithm in step S6 are:

[0168] B1, parameter initialization, the number of iterations of the drone’s position and resource allocation initialization;

[0169] B2, the cycle starts;

[0170] B3. Given, use Algorithm 1 to solve the user association and drone location problems and obtain the optimal solution for the iteration;

[0171] B4. Given {A (t+1) , V (t+1)}, use CVX to solve the UAV transmission power problem and get the optimal solution P for the t+1th iteration (t+1) ;

[0172] B5, t = t + 1, calculate {A (t+1) , P (t+1) , V (t+1)}, until the relative increase of the minimum rate of all users is lower than the threshold λ or t>T;

[0173] B6. Get the user association A, drone position V and transmission power P.

[0174] In the step S5, a first-order Taylor expansion and an approximate method are used to transform a non-convex problem into a convex problem, and a standard convex optimization tool is used to solve the convex problem.

[0175] In step S2, the multi-UAV location deployment and resource allocation scenario is modeled. The simulation experiment area is 4km*4km, the number of UAVs is set to [4, 5], and the number of ground users is set to [25, 30]. Compared with the "'User matching and spectrum resource joint optimization method in multi-UAV assisted communication' mentioned in the most similar implementation scheme", the present invention can achieve full coverage of ground users, and takes into account the optimization of UAV transmission power. It is proposed to decompose the optimization problem into two sub-problems based on the block coordinate descent method to speed up the convergence of the algorithm. Ground user deployment is as follows: Figure 7 The simulation experiment obtained the coverage of 4 drones and 5 drones for all ground users, as shown in Figure 8 and Fig. 9 shown.

[0176] In step S4, the user association and horizontal position of the drone are obtained according to the idea of ​​solving the problem of drone user association and resource allocation by the ant lion algorithm, and then the height of the drone is obtained according to the formula. Compared with the "'User matching and spectrum resource joint optimization method in multi-drone assisted communication' mentioned in the most similar implementation scheme", the present invention can quickly obtain the specific position of the drone, thereby realizing the deployment of multiple drones. The simulation experiment obtained the iteration graph of the ant lion algorithm. Fig.10 shown.

[0177] In step S6, the iterative optimization algorithm is an iterative algorithm based on the block coordinate descent method. The number of iterations of the algorithm changes with the transmission rate of the ground users until convergence is achieved. Therefore, compared with the "'User matching and spectrum resource joint optimization method in multi-UAV assisted communication' mentioned in the most similar implementation scheme", the overall algorithm based on the block coordinate descent method proposed in the present invention has fast algorithm convergence, can reduce algorithm complexity, and maximize the rate of all ground users. The iteration diagram of the overall algorithm is shown in Figure 2. Fig.11 shown.

[0178] In the overall algorithm in step S6, the simulation experiment obtained the change of the user transmission rate under different algorithms with the number of drones and the number of ground users. Compared with the "Joint optimization method of user matching and spectrum resources in multi-drone assisted communication" mentioned in the "most similar implementation scheme", as the number of drones increases, the minimum values ​​of all user rates under the proposed algorithm, k-means algorithm, greedy algorithm and random algorithm increase slowly. The reason is that when the number of drones is small, more transmission power is needed to improve the communication quality of users at the edge of the considered area. On the one hand, when the number of drones is large, they can be deployed closer to the ground users, and the channel gain between them can be improved, thereby reducing the transmission power of the drones. On the other hand, the more drones there are, the more communication resources can be provided. Both aspects will help to improve the minimum user rate. Therefore, the more drones there are, the larger the minimum user rate. It can be seen that the minimum user rate of the algorithm proposed in this paper is higher than that of the other three algorithms under different numbers of drones. As the number of ground users increases, the minimum value of the user transmission rate gradually decreases. The reason is that when the transmission power of the drones is the same and when the number of ground users is small, each user can share a larger bandwidth. When the number of users increases and the network resource supply becomes tight, the amount of resources available to each user changes relatively little. Therefore, it is not difficult to see that the minimum user rate of the algorithm proposed in this paper is higher than that of the other three algorithms under different numbers of ground users. The user transmission rate changes with the number of drones as shown in the figure Fig.12 The user transmission rate changes with the number of users as shown in Fig.13 As shown in the figure, the user transmission rate changes with the number of users and the number of drones. Fig.14 shown.

[0179] Although the specific embodiments of the present invention are described above, it should be understood by those skilled in the art that this is only for illustration and the protection scope of the present invention is defined by the appended claims. Those skilled in the art may make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, but these changes and modifications all fall within the protection scope of the present invention.

Claims

1. A location deployment and resource allocation method for multi-UAV communication, characterized by: The method comprises the following steps: S1, parameter and model initialization; S2, modeling, modeling the UAV location deployment and resource allocation problems; S3, block coordinate descent method to split the problem, using block coordinate descent method to split the mixed integer non-convex problem; S4, user association and drone 3D position, the ant lion algorithm obtains the drone user association and 3D position; S5, UAV transmission power, using successive convex approximation technology to optimize the UAV transmission power; S6, iterative optimization algorithm to obtain a suboptimal solution.

2. The method for position deployment and resource allocation for multi-UAV communication according to claim 1, characterized in that: In the step S1, the drone acts as a temporary base station and serves all ground users within the coverage area. One drone can cover multiple ground users, and the drone establishes communication with ground users through a downlink.

3. The method for position deployment and resource allocation for multi-UAV communication according to claim 1, characterized in that: The models established in step S2 are a channel model, a communication model and an optimization model.

4. The method for position deployment and resource allocation for multi-UAV communication as claimed in claim 3, characterized in that: The air-to-ground link in the channel model adopts a hybrid channel model, and the hybrid channel model is a line-of-sight link and a non-line-of-sight link.

5. The method for position deployment and resource allocation for multi-UAV communication according to claim 1, characterized in that: In the S3 step, the mixed integer non-convex problem is split into two sub-problems, which are the user association and UAV two-dimensional deployment problem and the UAV transmission power problem.

6. The method for position deployment and resource allocation for multi-UAV communication according to claim 1, characterized in that: The specific operation method of the S4 step is: Step 1: User association and two-dimensional deployment of drones; Step 2: Determine the altitude of the drone; Step 3: Jointly optimize user association and drone location deployment.

7. The method for position deployment and resource allocation for multi-UAV communication according to claim 6, characterized in that: In step 3, the improved ant lion algorithm is used for optimization, and the specific operation process is as follows: A1. Start; A2. Initialize various parameters; A3. Randomly generate population positions, calculate antlion fitness, and select elite antlions; A4. Select antlions through a roulette mechanism; A5, ants randomly walk around the ant lion; A6. Update the ant position; A7. Calculate the objective function value and select the antlion with the highest fitness as the elite antlion; A8, determine whether the number of iterations reaches the upper limit, if yes, execute step A9, if no, execute step A4; A9. End.

8. The method for position deployment and resource allocation for multi-UAV communication according to claim 1, characterized in that: In the step S5, the SCA technology is used to optimize the transmission power of the UAV.

9. The method for position deployment and resource allocation for multi-UAV communication according to claim 1, characterized in that: The specific operation steps of the iterative optimization algorithm in step S6 are: B1, parameter initialization, the number of iterations of the drone’s position and resource allocation initialization; B2, the cycle starts; B3. Given, use Algorithm 1 to solve the user association and drone location problems and obtain the optimal solution for the iteration; B4. Given {A (t+1) ,V (t+1) }, use CVX to solve the UAV transmission power problem and get the optimal solution P for the t+1th iteration (t +1) ; B5, t = t + 1, calculate {A (t+1) ,P (t+1) , V (t+1) }, until the relative increase of the minimum rate of all users is lower than the threshold λ or t>T; B6. Get the user association A, drone position V and transmission power P.

10. The method for position deployment and resource allocation for multi-UAV communication according to claim 1, characterized in that: In the step S5, a first-order Taylor expansion and an approximate method are used to transform a non-convex problem into a convex problem, and a standard convex optimization tool is used to solve the convex problem.

Images