Dual UAV-assisted ultra-low power consumption data acquisition method for green stereoscopic coverage

Abstract

The present invention provides a dual-UAV assisted ultra-low power consumption data acquisition method for green stereoscopic coverage, which belongs to the field of resource optimization of UAV uplink communication. The UAVs communicate by periodically flying above the user according to the designed circular trajectory, and the trajectories of the two UAVs differ by half a cycle. In order to make more effective use of these limited resources, we use device-to-device (D2D) technology to enable direct communication between devices. In each D2D cluster, the device with the least total hop count is selected as the central node to communicate with the UAV. On the basis of meeting the user's service quality, optimize resource allocation to maximize the energy efficiency of the system. Based on this model, the present invention is a design method for combining UAV trajectories, user scheduling and power allocation, which can meet the user's communication requirements and maximize the energy efficiency of the system.

Description

Technical Field

[0001] The present invention belongs to the technical field of resource allocation optimization of unmanned aerial vehicle multi-device wireless communication networks, and relates to a joint allocation strategy for power, unmanned aerial vehicle trajectory and device scheduling of a dual-unmanned aerial vehicle communication network with wide-area multi-devices, and specifically to a dual-unmanned aerial vehicle assisted ultra-low power consumption data acquisition method for green stereoscopic coverage. Background Art

[0002] UAV communication has attracted much attention as an important supplement to current ground communication due to its flexible configuration, high mobility, low cost and portability. When the ground network fails partially or completely, UAVs can provide emergency communication services. In fact, UAV-assisted wireless communication is a very promising solution that can provide wireless connections for users who lack infrastructure or whose infrastructure is damaged. At the same time, the air-to-ground communication channel of the UAV can be approximated as a high-quality line-of-sight link (LoS). The UAV can act as a base station or relay to provide short-term high-speed wireless connections for ground users. In addition, compared with ground base stations, the flexible location deployment of UAVs can further improve the performance of the system. Therefore, UAVs have great development potential in future communication systems.

[0003] An important application of drone communication is data collection and condition monitoring. In forest scenarios, due to the obstruction of tall trees, ground base stations need to be built relatively tall and have a relatively small coverage area, so more base stations need to be built. In addition, the post-maintenance of infrastructure such as base stations is also difficult, so it costs a lot to build ground base stations in forest scenarios. At this time, drone base stations are a very good choice. Compared with ground base stations, drone base stations do not require a lot of infrastructure, have very low maintenance costs, and can cover a larger area due to their high mobility. In addition, since they can establish LoS links, their power consumption is relatively small.

[0004] However, the energy that a drone can carry is very limited, and since drones work in the air, it is difficult to recharge them with energy. Therefore, it is very important to reasonably configure these system resources to utilize these energies more efficiently. However, this resource allocation problem has always been a difficult problem. This is because the variables in this problem are generally coupled with each other and are usually non-convex, so it is difficult to solve directly. Based on the above problems, the present invention considers a dual-drone communication network with wide-area multi-users, and maximizes the energy efficiency of the system by optimizing the resources in the system. Summary of the invention

[0005] In view of the above problems existing in the prior art, the present invention provides a resource allocation strategy to maximize the energy efficiency of the system. Two drones are deployed in the air as base stations to serve K evenly distributed devices. The drones communicate by periodically flying over the devices according to the designed circular trajectory. The trajectories of the two drones differ by half a cycle, which not only ensures low latency, but also ensures that the ground equipment can maintain high transmission rate communication more stably. In order to make more effective use of these limited resources, the present invention uses device-to-device (D2D) technology to enable direct communication between devices. In each D2D cluster, the device with the least total hop count is selected as the central node to communicate with the drone. On the basis of meeting the quality of service of the equipment, the drone trajectory, device scheduling and power allocation are jointly optimized to maximize the energy efficiency of the system.

[0006] In order to achieve the above objectives, the technical solution adopted by the present invention to solve the technical problem is as follows:

[0007] A dual-UAV assisted ultra-low power consumption data collection method for green stereoscopic coverage includes the following steps:

[0008] The first step is to build a system model:

[0009] 1) If Figure 1 , two drones are deployed in the air as base stations, and through device scheduling, they jointly serve K devices evenly distributed in an area with a radius of R. Assume that the working altitude of the two drones is H and the flight period is T. The trajectories of the two drones differ by half a period. When the device distance is less than the threshold d l When , D2D connections can be established between devices, and due to power constraints, the number of devices in a D2D cluster cannot exceed 10. In each D2D cluster, the one with the least total hops is selected as the central node to communicate with the drone. Assume that the set Λ={1,2,…,M} represents all M communication nodes, including D2D central nodes and devices that cannot establish D2D connections. On the basis of meeting the device service quality, the drone trajectory, device scheduling and power allocation are jointly optimized to maximize the energy efficiency of the system.

[0010] 2) In the system model, bold letters represent matrices or sets, and non-bold letters represent scalars; Represents a matrix space with dimension 2×1.

[0011] 3) U m Represents the mth communication node, the flight period T of the drone is discretized into N time slots, each time slot lasts for τ, and we can get:

[0012]

[0013] Assume Um The horizontal position is The position of the i-th UAV in the n-th time slot is Define its set as W = {w i [n]|i∈{1,2},1≤n≤N+1}. Therefore, in the nth time slot, U m The distance to the i-th UAV can be expressed as

[0014]

[0015] Since the UAV air-to-ground channel can be regarded as a LoS link, in the nth time slot, U m The channel coefficient with the i-th UAV |h i,m,n | 2 It can be expressed as:

[0016]

[0017] Among them, ρ 0 Represents the channel gain per unit distance.

[0018] Since the power of direct communication between devices is very small and the distance from the drone is relatively far, the interference caused by it can be ignored. Assuming that the drone serves the transmission node based on frequency division multiple access technology, in the nth time slot, device U m The uplink transmission rate between the i-th UAV can be expressed as

[0019]

[0020] Where P m,n For U m The transmission power in the nth time slot is defined as P = {P m,n |m∈Λ,1≤n≤N} is the set of transmission power, σ 2 Represents the noise power of high-speed white noise.

[0021] At the same time, the present invention introduces a binary variable α i,m,n , and use α={α i,m,n |i∈{1,2},m∈Λ,1≤n≤N} represents a set of binary variables. When α i,m,n =1, indicating that in the nth time slot, U m Establish a connection with the i-th drone, otherwise it means no connection is established. In order to ensure that each device can establish a connection with and only with one drone in each time slot, it needs to meet the following conditions:

[0022] α 1,m,n +α 2,m,n =1,m∈Λ,1≤n≤N. (5)

[0023] Therefore, in the nth time slot, device U m The uplink transmission rate can be expressed as:

[0024] R m,n =α 1,m,n R 1,m,n +α 2,m,n R 2,m,n . (6)

[0025] Define that in the nth time slot, the speed of the i-th UAV is For the convenience of description, assume that the set is V = {v i [n]|i∈{1,2},1≤n≤N+1}. Since the drone flies in a periodic circular trajectory, the speed of the drone needs to satisfy v i [1] = v i [N+1], so in the nth time slot, the upper bound of the propulsion power of the two drones can be calculated as:

[0026]

[0027] where c 1 and c 2 Indicates two parameters related to the mass, wing area, air weight, etc. of the drone. g indicates the standard number is 9.8m / s 2 The gravitational acceleration, is the acceleration of the ith UAV in the nth time slot, and its set can be expressed as A = {a i [n]|i∈{1,2},1≤n≤N+1}.

[0028] Definition P c is the circuit loss of each drone. Therefore, the energy efficiency of the system can be calculated as:

[0029]

[0030] In the second step, according to the specific settings of the first step, determine the objective function and list the optimization problem:

[0031] According to the derivation of the first step, on the basis of meeting the service quality requirements of the equipment, the energy efficiency maximization problem of the system can be expressed as:

[0032]

[0033] R m,n ≥η m , (9b)

[0034] α 1,m,n +α 2,m,n =1,α i,m,n∈{0,1},i∈{1,2}, (9c)

[0035]

[0036]

[0037] w i [N+1]=w i [1],v i [N+1]=v i [1],a i [N+1]=a i [1],(9f)

[0038]

[0039] v i [n+1]-v i [n] = a i [n]τ, (9h)

[0040]

[0041]

[0042] Among them, P max is the maximum transmission power of the ground equipment, η m is the service quality requirement of the transmission node, v min , v max and a max They represent the minimum speed, maximum speed and maximum acceleration of the drone respectively.

[0043] The third step is to design an algorithm to solve the optimization problem:

[0044] Problem (9) is a non-convex mixed integer optimization problem and the variables are coupled with each other, so it is difficult to solve directly. The present invention proposes a solution to simplify problem (9). First, an algorithm for establishing D2D connection is proposed to enable devices to communicate directly. Finally, an alternating iterative algorithm is proposed to optimize device scheduling, drone trajectory and transmission power to maximize energy efficiency.

[0045] 1) D2D connection algorithm

[0046] In order to cover more devices, the present invention first proposes an algorithm for establishing a D2D connection, assuming that when the distance between the devices is less than the distance threshold d l In order to establish a connection, due to power constraints, the number of devices in each D2D cluster does not exceed 10. This algorithm can be used to divide devices into different D2D clusters and select the central node. The principle of the algorithm is as follows:

[0047] First, all devices are clustered by D2D, using two loops. The outer loop is used to select devices that need to be D2D, and the inner loop is used to assign devices that can be connected to them to a D2D cluster. The specific steps are as follows: Assume that the current set of all ground devices that are not D2D clustered is S K , Initialization: S K ={1,2,…,K}, the number of outer iterations e = 1, assuming K e For S K The current number of devices in S e represents the D2D cluster in the e-th iteration.

[0048] 1.1) When e is less than or equal to K e , do the following:

[0049] 1.2) Initialization: Current D2D cluster S e ={S K The number of devices in the current cluster is D e =1, the number of inner layer iterations j=1.

[0050] 1.3) When j is less than or equal to D e , do the following:

[0051] 1.4) Calculate S e The jth device in S K The distance of the remaining devices in the , assuming the distance is less than d l The number of devices is e j .

[0052] 1.5) Determine e j +D e Is it less than 10?

[0053] 1.6) When it is less than 10, j Devices are assigned to S e and update j=j+1, D e =D e +e j .

[0054] 1.7) When it is greater than or equal to 10, the first 10-D e Devices are assigned to S e and update D e =10, ending the current loop.

[0055] Repeat 1.3)-1.7) until the current cycle ends

[0056] 1.8) S e Device from SK Remove it, update e=e+1, and calculate the current S K The number of elements in K e .

[0057] Repeat 1.1) to 1.8) until all ground devices are allocated, and the number of D2D clusters is e-1. In order to save energy, the present invention selects the device with the least total hop count as the central node, and then performs the following operations to select the central node: Initialization: Iteration number l=1.

[0058] 1.9) When l is less than e, perform the following operations:

[0059] 1.10) Initialization: number of iterations h = 1, minimum total number of hops f min = 100. Calculate the lth D2D cluster S l The number of devices in D l When h is less than or equal to D l , do the following:

[0060] 1.11) Calculate the number of other devices in the cluster to S l The total number of hops f to the hth device in .

[0061] 1.12) Determine whether f is less than f min .

[0062] 1.13) If f <f min , then update f min =f, and update the current central node to the device, h = h + 1. Otherwise, do nothing else, just update h = h + 1.

[0063] Through operations 1.9)-1.13), the central nodes in all D2D clusters can be selected and the total number of hops can be calculated.

[0064] 2) Alternately iterate and optimize equipment scheduling, drone trajectory and transmission power

[0065] Through the above method, ground devices are assigned to different D2D clusters. In order to solve problem (9), the present invention decomposes it into three sub-problems, optimizing device scheduling, UAV trajectory and transmission power separately, and finally proposes an iterative algorithm to alternately optimize the three sub-problems to solve (9).

[0066] 2.1) Equipment scheduling optimization

[0067] Fixing the drone trajectory and transmission power and optimizing the equipment scheduling separately, problem (9) can be decomposed into:

[0068]

[0069] α1,m,n +α 2,m,n =1, (10b)

[0070] α i,m,n ∈{0,1},i∈{1,2}. (10c)

[0071] This is a mixed integer non-convex optimization problem and cannot be solved directly using the CVX toolbox in MATLAB. Therefore, the present invention converts the binary variable into a continuous variable. Its satisfaction At the same time, since the trajectory and transmission power of the UAV are fixed, its objective function can be converted to Through the above transformation, problem (10) can be transformed into a linear programming problem. At the same time, its optimal solution is equivalent to the optimal solution of problem (10) and can be directly solved using the CVX toolbox.

[0072] 2.2) Drone trajectory optimization

[0073] According to the optimized equipment scheduling and fixed transmission power, the UAV trajectory optimization problem can be expressed as:

[0074]

[0075]

[0076] (9e)-(9j), (11c)

[0077] Since the objective function, constraints (11a) and (11b) are all non-convex, this problem is a non-convex fractional optimization problem. Inspired by the Dinkelbach algorithm, we first use the first-order Taylor expansion to transform the numerator of the objective function into a concave function, and then introduce a slack variable t i [n], which satisfies The denominator of the objective function is converted into a convex function. At the same time, constraint (11a) is also converted into a convex constraint, and constraint (11b) is converted into three convex constraints using the first-order Taylor expansion. Finally, the Dinkelbach algorithm is used to convert the problem into a convex optimization problem that can be directly solved by CVX.

[0078] 2.3) Transmission power optimization

[0079] According to the results of optimized equipment scheduling and UAV trajectory, the transmission efficiency is optimized separately, and problem (9) can be reformulated as:

[0080]

[0081]

[0082] R m,n ≥η m , (12b)

[0083] This problem is a fractional optimization problem, and the numerator of the objective function is a concave function and the denominator is a convex function. Therefore, according to the Dinkelbach algorithm, the problem can be directly converted into a convex optimization problem and solved by the CVX toolbox.

[0084] 2.4) Design of alternating optimization algorithm

[0085] Through the above method, the three sub-problems have been transformed into convex optimization problems, which can all be solved using CVX. In order to solve (9), an algorithm is proposed to alternately optimize the three sub-problems to solve the original problem. The specific steps are as follows:

[0086] 2.4.1) Initialize device scheduling, drone trajectory and transmission power. r and s represent the number of inner and outer layer iterations. Set the initial value r = 0.

[0087] 2.4.2) Initialize s=0;

[0088] 2.4.3) Based on W s 、V s , A s and P r , solve problem (10) and get the optimized schedule α s+1 ;

[0089] 2.4.4) Based on α s+1 and P r , solve problem (11) and get the optimized trajectory W s+1 、V s+1 , A s+1 ;

[0090] 2.4.5) Update s = s + 1;

[0091] 2.4.6) Repeat 2.4.3)-2.4.5) until the device scheduling and drone trajectory converge.

[0092] 2.4.7) Update r = r + 1;

[0093] 2.4.8) Repeat 2.4.2)-2.4.7) until all convergence.

[0094] The above optimization algorithm can be used to solve the optimization problem (9) listed in the second step, and give the optimal transmission power, equipment scheduling and drone trajectory to maximize the system energy efficiency.

[0095] The beneficial effects of the present invention are:

[0096] The present invention considers a dual-UAV wide-area multi-device system, where two UAVs fly to serve ground devices, and proposes an optimization problem to maximize energy efficiency. A new solution is provided to optimize resource allocation in UAV communication. First, ground devices can be connected together through the proposed D2D algorithm to improve coverage performance. Through reasonable resource allocation, the communication requirements of the devices in the system are firstly guaranteed to be met, and the energy efficiency of the system can also be maximized. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Figure 1 This is a system model diagram with dual UAVs and wide-area multi-devices.

[0098] Figure 2 This is a D2D rendering with 2,000 ground devices.

[0099] Figure 3 This is a curve showing the energy efficiency of the system as the number of iterations increases.

[0100] Figure 4 This is a curve showing the energy efficiency of the system as the flight cycle of the drone increases.

[0101] Figure 5 Trajectory plot optimized for systems with 2000 ground devices.

[0102] Figure 6 This is a comparison chart of the energy efficiency of the system under different schemes. DETAILED DESCRIPTION

[0103] In order to better understand the above technical solution, a specific analysis is given below in conjunction with the accompanying drawings and specific implementation methods.

[0104] The first step is to build a system model:

[0105] 1) If Figure 1 For the system model, two drones are deployed in the air as base stations to serve K = 2000 devices evenly distributed in an area with a radius of R = 8000m. Assume that the working altitude of the drone is H = 3000m and the flight period is T = 750s. The trajectories of the two drones differ by half a period. When the device distance is less than the threshold d l =400m, D2D connections can be established between devices, and due to power constraints, there cannot be more than 10 devices in a D2D cluster. In each D2D cluster, the one with the least total hops is selected as the central node to communicate with the drone. Assume that the set Λ={1,2,…,M} represents all M communication nodes, including the central node and devices that cannot establish D2D connections. On the basis of meeting the device service quality, the drone trajectory, device scheduling and power allocation are jointly optimized to maximize the energy efficiency of the system.

[0106] In the second step, based on the specific settings of the first step and the inferences of formulas (1) to (8), the objective function is determined and the optimization problem is listed as follows:

[0107] Parameters of the drone c 1 and c 2 They are set to 1.8×10 -3 and 4500, the circuit loss of both drones is 160W. Set the reference channel gain ρ 0 is -46dB, the noise power σ 2 is -118dBm. According to the derivation of the first step, on the basis of meeting the service quality requirements of the equipment, the energy efficiency maximization problem of the system can be expressed as:

[0108]

[0109] R m,n ≥η m , (13b)

[0110] α 1,m,n +α 2,m,n =1,α i,m,n ∈{0,1},i∈{1,2}, (13c)

[0111]

[0112]

[0113] w i [N+1]=w i [1],v i [N+1]=v i [1],a i [N+1]=a i [1], (13f)

[0114]

[0115] v i [n+1]-v i [n] = a i [n]τ, (13h)

[0116]

[0117]

[0118] Where P max =0.6W is the maximum transmission power of the ground equipment, η m =1bit / s / Hz is the service quality requirement of the transmission node, vmin =10m / s,v max =100m / s and a max =5m / s 2 They represent the minimum speed, maximum speed and maximum acceleration of the drone respectively.

[0119] The third step is to design an algorithm to solve the optimization problem:

[0120] Problem (13) is a non-convex mixed integer optimization problem and the variables are coupled with each other, so it is difficult to solve directly. The present invention proposes a solution to simplify problem (13). First, an algorithm for establishing D2D connection is proposed to enable devices to communicate directly when connected. Finally, an alternating iterative algorithm is proposed to optimize device scheduling, drone trajectory and transmission power to maximize energy efficiency.

[0121] 1) D2D connection algorithm

[0122] In order to cover more devices, the present invention first proposes an algorithm for establishing a D2D connection, assuming that when the distance between the devices is less than the distance threshold d l = 400m to establish a connection. In addition, due to power constraints, the number of devices in each D2D cluster does not exceed 10. This algorithm can be used to divide devices into different D2D clusters and select the central node.

[0123] The results are as follows Figure 2 , repeating 1.1) to 1.8) can divide 2000 ground devices into 314 D2D clusters, that is, the number of communication nodes M = 314. In order to save energy, the present invention selects the device with the least total hop count as the central node, and through operations 1.9)-1.13) the central nodes in all D2D clusters can be selected and the total hop count can be calculated.

[0124] 2) Alternately iterate and optimize equipment scheduling, drone trajectory and transmission power

[0125] Through the above method, ground devices are assigned to different D2D clusters. In order to solve problem (13), the present invention decomposes it into three sub-problems, optimizing device scheduling, UAV trajectory and transmission power separately, and finally proposes an iterative algorithm to alternately optimize the three sub-problems to solve (13).

[0126] 2.1) Equipment scheduling optimization

[0127] Fixing the UAV trajectory and transmission power and optimizing the equipment scheduling separately, problem (13) can be decomposed into

[0128]

[0129] α 1,m,n+α 2,m,n =1, (14b)

[0130] α i,m,n ∈{0,1},i∈{1,2}. (14c)

[0131] This is a mixed integer optimization problem and cannot be solved directly using the CVX toolbox in MATLAB. Therefore, the present invention converts the binary variable into a continuous variable Its satisfaction At the same time, since the trajectory and transmission power of the UAV are fixed, its objective function can be converted to Through the above transformation, problem (14) can be transformed into a linear programming problem. At the same time, its optimal solution is equivalent to the optimal solution of problem (14) and can be directly solved using the CVX toolbox.

[0132] 2.2) Drone trajectory optimization

[0133] According to the optimized equipment scheduling and fixed transmission power, the UAV trajectory optimization problem can be expressed as:

[0134]

[0135]

[0136] (13e)-(13j), (15c)

[0137] Since the objective function, constraints (15a) and (15b) are all non-convex, this problem is a non-convex fractional optimization problem. Inspired by the Dinkelbach algorithm, we first use the first-order Taylor expansion to transform the numerator of the objective function into a concave function, and then introduce a slack variable t i [n], which satisfies The denominator of the objective function is converted into a convex function. At the same time, constraint (15a) is also converted into a convex constraint, and constraint (15b) is converted into three convex constraints using the first-order Taylor expansion. Finally, the Dinkelbach algorithm is used to convert the problem into a convex optimization problem that can be directly solved by CVX.

[0138] 2.3) Transmission power optimization

[0139] According to the results of optimized equipment scheduling and UAV trajectory, the transmission efficiency is optimized separately, and problem (13) can be reformulated as:

[0140]

[0141] R m,n ≥η m , (16b)

[0142] This problem is a fractional optimization problem, and the numerator of the objective function is a concave function and the denominator is a convex function. Therefore, according to the Dinkelbach algorithm, the problem can be directly converted into a convex optimization problem and solved by the CVX toolbox.

[0143] 2.4) Design of alternating optimization algorithm

[0144] Through the above method, the three sub-problems have been transformed into convex optimization problems, which can all be solved using CVX. To solve (13), execute 2.4.1)-2.4.8), and the results are as follows:

[0145] Figure 3 The energy efficiency curve of the system is shown as the number of iterations increases. It can be seen from the figure that the solution proposed by the present invention is convergent and can effectively improve the energy efficiency of the system.

[0146] Figure 4 The energy efficiency curve of the system is shown as the flight cycle of the drone increases. Figure 4 It can be seen that as time goes by, the energy efficiency shows a trend of increasing first and then decreasing. This is because the longer the flight cycle, the larger the radius of the circular trajectory of the drone's flight. When the drone is closer to the center of the ground equipment, the equipment in the edge area requires more power, which will result in lower energy efficiency. Similarly, when the drone is close to the edge area equipment, the center area requires more power, which will also result in lower energy efficiency. Therefore, according to Figure 4 It can be determined that the time it takes for the drone to fly one circle is 750s. It can also be seen from the figure that the energy efficiency increases with the maximum transmission power P max The increase of P shows a trend of increasing first and then remaining constant. max The optimal power P is not achieved * , so the energy efficiency will increase with P max increases with the increase of * Afterwards, P max The increase of has no effect on energy efficiency. Therefore, according to this figure, the maximum transmission power can be set to 0.6W.

[0147] Figure 5 The trajectory diagram optimized for a system with 2000 ground devices. It can be seen from the figure that the proposed scheme can effectively optimize the system resource configuration. Since the two drones fly in opposite directions, their flight trajectories differ by half a cycle, so the trajectories between the two drones in the figure are overlapping. The above simulation results can verify the effectiveness of this scheme.

[0148] In order to highlight the superiority of this solution, the present invention sets up several different comparison schemes. This solution is Scheme 1. In Schemes 2 and 3, the trajectory of the drone is not optimized, and it is set to a circular trajectory with a radius of 2500m and 2000m respectively. Scheme 4 is a dual ground base station solution, with two base stations located at and , and because of its low height, the Rayleigh channel is used to simulate the wireless transmission channel, and its corresponding path loss coefficient is set to 2.2. It should be noted that, unless otherwise specified, the simulation parameters all use the parameters in the above steps.

[0149] Figure 6 This is a comparison chart of the energy efficiency of the system under different schemes. It can be seen that this scheme is better than the other three schemes. It can achieve higher energy efficiency at the same power, and only a smaller maximum transmission power is needed to achieve the optimal energy efficiency. The superiority of this scheme can be seen from the comparison results.

[0150] The above-described embodiments merely express the implementation methods of the present invention, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A dual-UAV assisted ultra-low power consumption data collection method for green stereoscopic coverage, characterized in that: The following steps are involved: The first step is to build a system model: 1) Two drones are deployed in the air as base stations, and through device scheduling, they jointly serve K devices evenly distributed in an area with a radius of R. Assume that the working altitude of the two drones is H and the flight period is T. The trajectories of the two drones differ by half a period. When the device distance is less than the threshold d l When , D2D connections can be established between devices, and the number of devices in a D2D cluster cannot exceed 10; In each D2D cluster, the one with the least total hop count is selected as the central node to communicate with the drones; Assume that the set Λ={1,2,…,M} represents all M communication nodes, including D2D central nodes and devices that cannot establish D2D connections; on the basis of satisfying the device service quality, jointly optimize the drone trajectory, device scheduling and power allocation to maximize the energy efficiency of the system; 2) In the system model, bold letters represent matrices or sets, and non-bold letters represent scalars; Represents a matrix space with a dimension of 2×1; 3) U m Represents the mth communication node, the flight period T of the drone is discretized into N time slots, each time slot lasts for τ, and we can get: Assume U m The horizontal position is The position of the i-th UAV in the n-th time slot is Define its set as W = {w i [n]|i∈{1,2},1≤n≤N+1}; Therefore, in the nth time slot, U m The distance to the i-th drone can be expressed as: In the nth time slot, U m The channel coefficient with the i-th UAV |h i,m,n | 2 It is expressed as: Where ρ0 represents the channel gain at unit distance; Assuming that the drone serves the transmission node based on frequency division multiple access technology, in the nth time slot, device U m The uplink transmission rate between the i-th UAV is expressed as: Among them, P m,n For U m The transmission power in the nth time slot is defined as P = {P m,n |m∈Λ,1≤n≤N} is the set of transmission power, σ 2 represents the noise power of high-speed white noise; At the same time, a binary variable α is introduced i,m,n , and use α={α i,m,n |i∈{1,2},m∈Λ,1≤n≤N} represents a set of binary variables; when α i,m,n =1, indicating that in the nth time slot, U m Establish a connection with the i-th drone, otherwise it means no connection is established; in order to ensure that each device can establish a connection with and only with one drone in each time slot, it needs to meet the following conditions: a 1,m,n +a 2,m,n =1,m∈Λ,1≤n≤N. (5) Therefore, in the nth time slot, device U m The uplink transmission rate is expressed as: R m,n =a 1,m,n R 1,m,n +a 2,m,n R 2,m,n . (6) Define that in the nth time slot, the speed of the i-th UAV is For the convenience of description, assume that the set is V = {v i [n]|i∈{1,2},1≤n≤N+1}; Since the drone flies in a periodic circular trajectory, the speed of the drone needs to satisfy v i [1] = v i [N+1], so in the nth time slot, the upper bound of the propulsion power of the two drones is calculated as: Among them, c1 and c2 represent two parameters related to the drone, and g represents the standard number of 9.8m / s 2 The gravitational acceleration, is the acceleration of the ith UAV in the nth time slot, and its set can be expressed as A = {a i [n]|i∈{1,2},1≤n≤N+1}; Define P c is the circuit loss of each drone; therefore the energy efficiency of the system can be calculated as: In the second step, according to the specific settings of the first step, determine the objective function and list the optimization problem: According to the derivation of the first step, on the basis of meeting the service quality requirements of the equipment, the energy efficiency maximization problem of the system is defined as problem (9), which can be expressed as: Question (9): s.t.0≤P m,n ≤P max ,m∈Λ,1≤n≤N,(9a) R m,n ≥η m ,(9b) a 1,m,n +a 2,m,n =1,a i,m,n ∈{0,1},i∈{1,2},(9c) w i [N+1]=w i [1],v i [N+1]=v i [1],a i [N+1]=a i [1],(9f) V i [n+1]-v i [n]=a i [n]τ,(9h) Among them, P max is the maximum transmission power of the ground equipment, η m is the service quality requirement of the transmission node, v min , v max and a max Respectively represent the minimum speed, maximum speed and maximum acceleration of the drone; The third step is to design an algorithm to solve the optimization problem: The proposed solution simplifies the problem (9) proposed in the second step. First, an algorithm for establishing D2D connections is proposed to enable devices to communicate directly. Finally, an alternating iterative algorithm is proposed to optimize device scheduling, drone trajectories and transmission power to maximize energy efficiency. 3.1) D2D connection algorithm Assume that when the distance between devices is less than the distance threshold d l In order to establish a connection, due to power constraints, the number of devices in each D2D cluster does not exceed 10. This algorithm can be used to divide devices into different D2D clusters and select the central node at the same time. 3.2) Alternative Iteration Optimization of Equipment Scheduling, UAV Trajectory and Transmission Power Through the above method, ground devices are assigned to different D2D clusters. In order to solve problem (9), it is decomposed into three sub-problems, defined as problem (10), problem (11), and problem (12), which are used to optimize device scheduling, UAV trajectory, and transmission power, respectively. Finally, an iterative algorithm is proposed to alternately optimize the three sub-problems to solve problem (9); 3.2.1) Equipment scheduling optimization; 3.2.2) UAV trajectory optimization; 3.2.3) Transmission power optimization; 3.2.4) Design an alternating optimization algorithm to solve the above problems. The specific steps are as follows: 3.2.4.1) Initialize device scheduling, drone trajectory and transmission power. r and s represent the number of inner and outer layer iterations. Set the initial value r = 0. 3.2.4.2) Initialize s=0; 3.2.4.3) Based on W s 、V s , A s and P r , solve problem (10) and get the optimized schedule α s+1 ; 3.2.4.4) Based on α s+1 and P r , solve problem (11) and get the optimized trajectory W s+1 、V s+1 , A s+1 ; 3.2.4.5) Update s = s + 1; 3.2.4.6) Repeat 3.2.4.3)-3.2.4.5) until the device scheduling and drone trajectory converge; 3.2.4.7) Based on the optimized scheduling and trajectory, solve problem (12) and obtain the optimized transmission power allocation P r+1 ; 3.2.4.8) Update r = r + 1; 3.2.4.9) Repeat 3.2.4.2)-3.2.4.8) until all convergence; The above optimization algorithm can be used to solve the problem (9) listed in the second step, and give the optimal transmission power, equipment scheduling and drone trajectory to maximize the system energy efficiency.

2. The dual-UAV-assisted ultra-low power consumption data acquisition method for green stereoscopic coverage according to claim 1 is characterized in that: The step 3.1) is as follows: First, all devices are clustered by D2D, using two loops, where the outer loop is used to select devices that need to be D2D, and the inner loop is used to assign devices that can be connected to them to a D2D cluster. The specific steps are as follows: Assume that the current set of all ground devices that are not D2D clustered is S K , Initialization: S K ={1,2,…,K}, the number of outer iterations e = 1, assuming K e For S K The current number of devices in S e represents the D2D cluster in the e-th iteration; 3.1.1) When e is less than or equal to K e , do the following: 3.1.2) Initialization: Current D2D cluster S e ={S K The number of devices in the current cluster is D e =1, inner layer iteration number j=1; 3.1.3) When j is less than or equal to D e , do the following: 3.1.4) Calculate S e The jth device in S K The distance of the remaining devices in the , assuming the distance is less than d l The number of devices is e j ; 3.1.5) Determine e j +D e Is it less than 10? 3.1.6) If it is less than 10, j Devices are assigned to S e and update j=j+1, D e =D e +e j ; 3.1.7) When the value is greater than or equal to 10, the first 10-D e Devices are assigned to S e and update D e =10, end the current loop; repeat 3.1.3)-3.1.7) until the current loop ends 3.1.8) S e The device is from S K Remove it, update e=e+1, and calculate the current S K The number of elements in K e ; Repeat 3.1.1) to 3.1.8) until all ground devices are allocated, then the number of D2D clusters is e-1; select the device with the least total hops as the central node, and then perform the following operations to select the central node: Initialization: number of iterations l = 1; 3.1.9) When l is less than e, perform the following operations: 3.1.10) Initialization: number of iterations h = 1, minimum total number of hops f min =100; calculate the lth D2D cluster S l The number of devices in D l ; When h is less than or equal to D l , do the following: 3.1.11) Calculate the number of other devices in the cluster to S l The total number of hops f of the hth device in; 3.1.12) Determine whether f is less than f min ; 3.1.13) If f <f min , then update f min =f, and update the current central node to the device, h=h+1; otherwise, do nothing else, just update h=h+1; Through operations 3.1.9)-3.1.13), the central nodes in all D2D clusters can be selected and the total number of hops can be calculated.

3. The dual-UAV-assisted ultra-low power consumption data acquisition method for green stereoscopic coverage according to claim 1 is characterized in that: In step 3.2, the specific steps for optimizing device scheduling, drone trajectory and transmission power separately are: 3.2.1) Equipment scheduling optimization Fixing the drone trajectory and transmission power and optimizing the equipment scheduling separately, problem (9) can be decomposed into: Question (10): stα 1,m,n R 1,m,n +a 2,m,n R 2,m,n ≥η m ,(10a) α 1,m,n +α 2,m,n ＝1,(10b) α i,m,n ∈{0,1},i∈{1,2}.(10c) Convert the binary variable into a continuous variable Its satisfaction At the same time, since the trajectory and transmission power of the UAV are fixed, its objective function can be converted to Through the above transformation, problem (10) can be transformed into a linear programming problem. At the same time, its optimal solution is equivalent to the optimal solution of problem (10), and can be directly solved using the CVX toolbox; 3.2.2) UAV trajectory optimization According to the optimized equipment scheduling and fixed transmission power, the UAV trajectory optimization problem can be expressed as: Question (11): s.t.R m,n ≥η m ,(11a) (9e)-(9j), (11c) First, the numerator of the objective function is converted into a concave function using the first-order Taylor expansion, and then a slack variable t is introduced. i [n], which satisfies The denominator of the objective function is converted into a convex function. At the same time, constraint (11a) is also converted into a convex constraint. Then, constraint (11b) is converted into three convex constraints using the first-order Taylor expansion. Finally, the Dinkelbach algorithm is used to convert the problem into a convex optimization problem that can be directly solved by CVX. 3.2.3) Transmission power optimization According to the results of optimized equipment scheduling and UAV trajectory, the transmission efficiency is optimized separately, and problem (9) can be reformulated as: Question (12): s.t.0≤P m,n ≤P max ,(12a) R m,n ≥η m ,(12b) According to the Dinkelbach algorithm, the problem can be directly converted into a convex optimization problem and solved using the CVX toolbox.

Images