An optimization method for full-duplex fixed-wing UAV relay system

Abstract

The present invention discloses an optimization method for a full-duplex fixed-wing UAV relay system with roll angle limitation. The method utilizes the location information of a source node and a destination node, optimizes and adjusts the flight radius, flight speed, and flight time of the UAV according to the amount of data that the source node needs to send to the destination node, the maximum and minimum flight speeds of the UAV, and the roll angle limitation. The method minimizes the UAV flight energy consumption while complying with the UAV flight speed and roll angle limitations and meeting the system's requirements for the amount of data sent.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to an optimization method for a full-duplex fixed-wing UAV relay system with roll angle limitation under a circular trajectory. Background Art

[0002] In recent years, UAV cooperative communication technology has become a hot topic in the field of wireless communications. Compared with traditional ground communications, UAV cooperative communication is easy to implement on-demand allocation, thus having higher communication efficiency; UAVs are highly mobile, making deployment more flexible and fast; the communication link between UAVs and ground terminals is more likely to be a line-of-sight channel than ground communications at the same communication distance, thus providing better channel transmission conditions. Therefore, UAVs will play an extremely important role in the future field of wireless communications. Their main applications include: (1) as temporary base stations; (2) as mobile relays; and (3) in the Internet of Things.

[0003] Currently, a large body of literature has studied the optimization of relay communication system capacity and spectral efficiency when drones are used as mobile relays. In reality, drones have limited onboard energy, and energy efficiency is considered a key issue in drone communications. Currently, most research on drone relay communications focuses on decode-and-forward relay communication, while research on amplify-and-forward relaying is limited. In particular, amplify-and-forward drone relay communications in full-duplex mode has not been considered and warrants further investigation. Furthermore, the flight radius and speed of fixed-wing drones operating in a circular trajectory are related to their roll angle, a problem often overlooked in the past and warrants further investigation. Summary of the Invention

[0004] The purpose of the present invention is to provide an optimization method for a full-duplex fixed-wing UAV relay system with roll angle limitation under a circular trajectory, so as to adjust the flight radius, flight speed and flight time of the UAV relay system, and minimize the UAV flight energy consumption while complying with the UAV flight speed limitation and roll angle limitation and meeting the system data transmission volume requirements, thereby improving the energy efficiency of the system.

[0005] The present invention is implemented as follows: a method for optimizing a full-duplex fixed-wing UAV relay system with roll angle limitation utilizes the location information of the source node and the destination node, optimizes and adjusts the flight radius, flight speed and flight time of the UAV according to the amount of data that the source node needs to send to the destination node and the maximum and minimum speeds of the UAV flight and the roll angle limitation, and minimizes the UAV flight energy consumption while complying with the UAV flight speed and roll angle limitations and meeting the system's requirements for the amount of data sent.

[0006] For a full-duplex fixed-wing UAV relay communication system with roll angle restriction under a circular trajectory, the source node S on the ground needs to send Q data volume to the destination node D. Assuming that the distance between S and D is L, since the distance between S and D is far, their direct link does not exist, and it is necessary to forward the data with the help of a fixed-wing UAV relay R flying at an altitude of H. The UAV relay R operates in full-duplex mode (receiving and transmitting signals simultaneously in the same frequency band) and adopts the Amplify-and-Forward (AF) relay protocol; the flight trajectory plane of the UAV relay R is parallel to the ground plane, and the projection of its flight trajectory on the ground plane is a circle with the midpoint of the line connecting the source node S and the destination node D as the center, r as the radius, and the circular flight speed as v; the UAV has a roll angle restriction during flight. ( The value is less than or equal to 45°), that is, the actual roll angle φ of the UAV flight needs to meet Assume that the wireless signal transmission channels from S to R and from R to D are mixed probabilistic channels, that is, mixed channels in which line-of-sight (LoS) and non-line-of-sight (NLoS) links exist probabilistically. Thus, at time t, the channel from S to R can be written as:

[0007]

[0008] where d SR (t) is the distance from S to R at time t, which can be specifically given by the following formula (2), θ is the wireless channel fading factor, represents the reference value of the channel gain at a distance of 1m, λ is the additional fading factor in the case of NLoS link, and P r SR -LOS (t) is the probability that the link between S and R is a LOS link at time t, which can be specifically given by formula (3), 1-P r SR-LOS (t) is the probability that the link between S and R is an NLOS link at time t;

[0009]

[0010]

[0011] It should be noted that: here we assume that the position of the source node S is (0,0,0), the position of the destination node D is (L,0,0), and the starting point of the UAV flight is In formula (3), arcsin(·) is the inverse sine function, α and β are the parameters of the mixed probability channel model, and the values ​​of α and β depend on the specific communication geographical environment;

[0012] Similarly, at time t, the channel from R to D can be written as:

[0013]

[0014] Among them, d RD (t) is the distance from R to D at time t, which can be specifically given by the following formula (5), P r RD-LOS (t) is the probability that the link between R and D is a LOS link at time t, which can be specifically given by the following formula (6): 1-P r RD-LOS (t) is the probability that the link between R and D is an NLOS link at time t;

[0015]

[0016]

[0017] Because of the above h SR (t) and h RD (t) are mixed probability channels of LoS and NLoS, which are difficult to process mathematically. Therefore, the probability of LoS and NLoS links is used to calculate h respectively. SR (t) and h RD The average value of (t) and Specifically, they can be given by formula (7) and formula (8);

[0018]

[0019]

[0020] It should be noted that: In formula (8) In the following, we will use the formula (7) and formula (8) and Replace h SR (t) and h RD (t) perform mathematical analysis;

[0021] Since the drone relay R works in full-duplex mode, the wireless signal transmitted by the drone's transmitting antenna will be received by the drone's receiving antenna. Therefore, the drone needs to use loop interference cancellation (LIC) technology to eliminate the loop interference generated in full-duplex mode. At time t, the signal received by the drone relay R is expressed as:

[0022]

[0023] Among them, PTX is the transmission power of S and R; x S is the transmitted signal of S (assuming the power is 1); x R is the transmitted signal of R (assuming the power is 1); h LI is the residual loop interference after LIC; R is the Gaussian white noise received by R (assuming the mean is 0 and the variance is σ 2 , that is, the average power of Gaussian white noise is σ 2 ), the drone will receive signal y R (t) is multiplied by the amplification factor ρ to become the transmitted signal, that is, x R =ρy R (t), here Assume that the drone relay R receives the signal y R If the processing of (t) is ideal enough without any delay, the signal received by node D is:

[0024]

[0025] where z D is the Gaussian white noise received by D (assuming the mean is 0 and the variance is σ 2 );

[0026] x R =ρy R Substitute (t) into formula (10) and perform mathematical operations to obtain

[0027]

[0028] The received signal-to-noise ratio of node D can be obtained from formula (11), which can be written as:

[0029]

[0030] in

[0031] In this way, the amount of data that the destination node can receive at time t is:

[0032]

[0033] According to the above analysis, the UAV energy consumption optimization problem can be written as

[0034]

[0035] stQ-Q D (T)≤0 (14b)

[0036] V min ≤v≤V max (14c)

[0037] φ≤ω (14d)

[0038] Here, T is the time to complete the forwarding of Q data volume, is the power consumption of a fixed-wing UAV flying a circular trajectory at a constant speed v and radius r, where g represents the acceleration due to gravity and c1 = ηC D0 B / 2, c2=2W 2 / [(πe0A R )ηB], η represents the air density, C D0 represents the zero lift drag coefficient, B represents the wing area, e0 is the wingspan efficiency, W represents the overall weight of the drone, A R represents the aspect ratio of the drone wing, V max and V min are the maximum and minimum flight speeds of the UAV, respectively. The constraint condition (14d) indicates that the actual roll angle φ of the UAV needs to meet the roll angle limit of the UAV, that is, the actual roll angle φ should be less than or equal to the maximum roll angle that the UAV can support. ( The value is less than or equal to 45°);

[0039] It should be noted that: the fixed-wing UAV's flight speed v and flight radius r satisfy the circular trajectory. where φ is the actual roll angle of the drone, and Need to be established constantly;

[0040] To understand the optimization problem (14), i.e., the optimization problem composed of equations (14a), (14b), (14c), and (14d), we start to simplify the constraint condition (14b); to do this, we first give log2[1+γ D (t)], which can be written as:

[0041]

[0042] Then according to the inequality And formula (15), we can further find log2[1+γ D (t)], which can be written as:

[0043]

[0044] when hour, P r SR-LOS (t) = P r RD-LOS (t), has a maximum value, so we can get log2[1+γ D(t)], which can be written as:

[0045]

[0046] It should be noted that: and It is only a function of r and has nothing to do with v;

[0047] According to equations (13) and (17), the lower bound of the amount of data that the destination node D can receive at time t is:

[0048]

[0049] in,

[0050]

[0051] Next, we use the formula (19) to obtain Q D The lower bound of (t) Instead of Q D (t), and transform φ≤ω Then the UAV energy consumption optimization problem, that is, problem (14) can be transformed into

[0052]

[0053]

[0054] V min ≤v≤V max (20c)

[0055]

[0056] It should be noted that: when the lower bound of the data that the destination node D can receive is greater than or equal to Q, the data that the destination node can actually receive will definitely be greater than or equal to Q;

[0057] To understand problem (20), i.e., the optimization problem composed of equations (20a), (20b), (20c), and (20d), first, simplify the constraints;

[0058] Observing formula (19), we can find that is an increasing function of t. Therefore, when the objective function (20a) of problem (20) takes the minimum value, (20b) will take the equal sign. Otherwise, the value of (20a) can be further reduced by shortening the time and satisfy the constraint condition. When (20b) takes the equal sign, there is

[0059]

[0060] It should be noted that T in formula (21) has nothing to do with v and is only an increasing function of r;

[0061] According to the relationship between the UAV's flight speed, flight radius and actual roll angle, we can get Then Substitute P W (r,v), we can get:

[0062]

[0063] For formula (22), find its partial derivative with respect to tanφ and set it to 0, and we can get

[0064]

[0065] That is to say, when formula (22) takes its minimum value, formula (23) holds;

[0066] From equations (20c) and (20d), we can get It always holds true; T given in formula (21) is an increasing function of r; let the UAV flight radius corresponding to the minimum value of formula (22) be r1, then the UAV flight radius r when the objective function formula (20a) of problem (20) takes the minimum value * Must be satisfied

[0067] To this end, we first use the one-dimensional search algorithm (reference: Jie Kexin, Han Lixing, Lin Youlian. Optimization Methods [M]. Tianjin: Tianjin University Press, 1997. 13-26.) to solve problem (24), that is, the optimization problem composed of equations (24a) and (24b). The one-dimensional search here is within the flight radius given by equation (24b), and the search obtains the UAV flight radius r2 when equation (24a) takes the minimum value;

[0068]

[0069]

[0070] Among them, Q is the amount of data sent from the source node S to the destination node D, r is the radius of the UAV flying in a circle at a constant speed v, It is the roll angle limit of the UAV during flight. Less than or equal to 45°, V min is the minimum flight speed of the UAV, e represents the natural constant, σ 2 represents the variance of Gaussian white noise, is the power consumption of a fixed-wing UAV flying a circular trajectory at a constant speed v and radius r, where g represents the acceleration due to gravity and c1 = ηC D0 B / 2, c2=2W 2 / [(πe0A R )ηB], where η represents the air density, C D0 represents the zero lift drag coefficient, B represents the wing area, e0 is the wingspan efficiency, W represents the overall weight of the drone, A R represents the aspect ratio of the drone wing, Among them, P TX is the transmit power of S and R, and are the average channel gains from S to R and from R to D at time t, respectively.

[0071] Then, substitute r2 as r into formula (23) to calculate the tangent value tanφ of the actual roll angle φ of the UAV at this time. If If true, calculate the corresponding flight speed at this time

[0072] like If true, then r2 is the optimal flight speed r of problem (20) * , the corresponding optimal flight speed is:

[0073]

[0074] like The optimal flight speed v of problem (20) is * =V min , the corresponding optimal flight radius r * It can be calculated by formula (26):

[0075]

[0076] like The optimal flight speed v of problem (20) is * =V max , the corresponding optimal flight radius r * It can be calculated by formula (27):

[0077]

[0078] If r2 is substituted into formula (23), the tangent value tanφ of the actual roll angle φ of the UAV does not satisfy Then let the actual roll angle φ of the drone be the maximum roll angle Then, the one-dimensional search algorithm is used to solve problem (28), that is, the optimization problem composed of equations (28a) and (28b). The one-dimensional search here is to perform a one-dimensional search on equation (28a) within the flight radius given by equation (28b) to obtain the corresponding UAV flight radius r3;

[0079]

[0080]

[0081] At this time, the flight speed corresponding to the flight radius r3 is Since r3 is obtained by searching within the range given by formula (28b), Established;

[0082] like If true, then r3 is the optimal flight radius r of problem (20) * , the optimal flight speed at this time is:

[0083]

[0084] like Then the optimal flight speed for problem (20) is v * =V max , the optimal flight radius at this time can be calculated by formula (30);

[0085]

[0086] Finally, the optimal flight radius r * Substituting r into formula (21) can calculate the optimal flight time T * .

[0087] The following is a method to solve the UAV energy consumption optimization problem, as follows:

[0088] Step 1: Use the one-dimensional search algorithm to solve problem (24) and obtain the UAV flight radius r2;

[0089] Step 2: Substitute r2 as r into equation (23) to obtain tanφ;

[0090] Step 3: If If true, calculate the corresponding flight speed at this time Otherwise jump to step 6;

[0091] Step 4: If If it holds, then the optimal flight radius r * =r2, optimal flight speed v * It can be calculated by formula (25), and then jump to step 9;

[0092] Step 5: If The optimal flight speed v * =V min , optimal flight radius r * It can be calculated by formula (26), and then jump to step 9;

[0093] Step 5: If If it holds, then the optimal flight speed v * =V max , optimal flight radius r * It can be calculated by formula (27), and then jump to step 9;

[0094] Step 6: If Then use the one-dimensional search algorithm to solve problem (28) and obtain the UAV flight radius r3;

[0095] Step 7: If If it holds, then the optimal flight radius r * =r3, optimal flight speed v * It can be calculated by formula (29), and then jump to step 9;

[0096] Step 8: If If true, the optimal flight speed is v * =V max , optimal flight radius r * It can be calculated by formula (30);

[0097] Step 9: Set the optimal flight radius r * Substitute r into formula (21) to calculate the optimal flight time T * ;

[0098] Step 10: The algorithm ends.

[0099] Compared to the shortcomings and deficiencies of existing technologies, the present invention offers the following advantages: It utilizes the location information of the source and destination nodes to optimize the drone's flight speed, flight radius, and flight time based on the drone's maximum and minimum flight speed limits, the drone's maximum roll angle limit, and the amount of data the source node needs to send to the destination node. This method minimizes the drone's flight energy consumption while meeting the system's data transmission requirements and maximum roll angle limits. Simulation experiments also demonstrate this optimization method's energy efficiency advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0100] Figure 1 Schematic diagram of the method of the present invention;

[0101] Figure 2 Comparison of energy consumption between the proposed optimization method and the minimum and maximum flight speeds of the UAV under different L values;

[0102] Figure 3 The actual amount of data received by the destination node using the proposed optimization method at the minimum and maximum flight speeds of the UAV under different L values;

[0103] Figure 4is the flight speed of the UAV using the proposed optimization method under different L values;

[0104] Figure 5 The actual roll angle of the UAV at the minimum and maximum flight speeds using the proposed optimization method under different L values;

[0105] Figure 6 Comparison of energy consumption between the proposed optimization method and the minimum and maximum flight speeds of the UAV under different Q values;

[0106] Figure 7 The actual amount of data received by the destination node using the proposed optimization method at different Q values ​​and the minimum and maximum flight speeds of the UAV;

[0107] Figure 8 is the flight speed of the UAV using the proposed optimization method under different Q values;

[0108] Figure 9 The actual roll angle of the UAV at the minimum and maximum flight speeds using the proposed optimization method under different Q values;

[0109] Figure 10 For different The energy consumption of the proposed optimization method is compared with that of the UAV at the minimum and maximum flight speeds under the given value;

[0110] Figure 11 For different The actual amount of data received by the destination node under the minimum and maximum flight speeds of the UAV using the proposed optimization method under the given value;

[0111] Figure 12 For different The flight speed of the UAV using the proposed optimization method under the value;

[0112] Figure 13 For different The actual roll angle of the UAV at the minimum and maximum flight speeds using the proposed optimization method under different values. DETAILED DESCRIPTION

[0113] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0114] like Figure 1As shown, the drone relay R always flies at a height of H, and the projection of its flight trajectory on the ground plane is a circle with the midpoint of the line connecting the source node S and the destination node as the center and r as the radius. The projection of the drone's starting point on the ground plane is the midpoint of the line connecting the source node S and the destination node, which is r-biased toward the direction of the S node. The flight line speed of the drone relay R is v. The drone R forwards the message to the destination node while receiving the signal sent by S. The drone receives the signal and forwards the message simultaneously in the same frequency band, that is, it works in full-duplex mode, and the relay forwarding adopts the AF protocol. Before the drone officially flies, it obtains r2 through a one-dimensional search for the optimization problem (24); it should be noted that when performing a one-dimensional search, the upper limit of r is originally infinite. In actual operation, a very large number can be set, such as L×10 9 ; Then, substitute r2 into formula (23) to calculate tanφ; Next, check whether tanφ is less than or equal to If tanφ is less than or equal to Then calculate the flight speed corresponding to r2 like Then r2 is the optimal flight radius r * , optimal flight speed v * It can be calculated by formula (25); if The optimal flight speed v * =V min , optimal flight radius r * It can be calculated by formula (26); if The optimal flight speed v * =V max , optimal flight radius r * It can be calculated by formula (27); if the tanφ calculated above is greater than Then let the drone follow the maximum roll angle To fly, we use the one-dimensional search algorithm to solve problem (28) and obtain the flight radius r3 of the drone. Then, we calculate the flight speed corresponding to r3. Check below Whether the flight speed limit is met; if The optimal flight radius r * =r3, optimal flight speed v * It can be calculated by formula (29); if The optimal flight speed is v * =V max , optimal flight radius r * It can be calculated by formula (30); finally, according to the obtained optimal flight radius r * Substitute r into formula (21) to calculate the optimal flight time T * .

[0115] For the optimization method proposed in this invention, this invention conducted simulation experiments and compared the system power consumption when the drone was flying at the minimum speed and the maximum speed. The experimental environment was Matlab environment. Assuming that the minimum and maximum speeds of the drone are v min =5m / s,v max =50m / s, the drone flies at a fixed height H=100m, and the signal transmission power of nodes S and R is P TX =-20dBm, the variance of the Gaussian white noise in the environment is σ 2 =-170dBm, channel gain per unit distance The channel fading factor θ = 3, the mixed probability channel model parameters α = 4.88, β = 0.43 (corresponding to the rural geographical environment), the additional fading factor λ = -10dB in the case of NLoS link, c1 = 9.26×10 -4 , c2=2250,|h LI | 2 =10 -6 .

[0116] Figure 2 、 Figure 3 、 Figure 4 and Figure 5 The energy consumption comparison of the optimization method and the minimum and maximum flight speed of the UAV under different L values, the actual amount of data received by the destination node, the flight speed of the UAV and the actual roll angle of the UAV are given respectively. Here, assuming Q = 100, The “V min ” and “V max " means that the UAV flies at the minimum and maximum flight speeds respectively, but the flight radius and flight time are optimized. "Proposed optimization algorithm" is the result obtained by using the proposed optimization method, that is, the result obtained by using the joint adjustment algorithm of the UAV flight speed, flight radius and flight time. "Simulation" means that the given values ​​are obtained by the interior point method in the Matlab optimization toolbox.

[0117] Figure 2 It is shown that the proposed optimization algorithm has performance advantages in terms of power consumption; Figure 3 It shows that the actual amount of data received by the destination node is greater than the set Q value, which shows the correctness of the proposed algorithm; Figure 4 It shows that the actual flight speed of the UAV meets the given constraints, which shows the correctness of the proposed algorithm; Figure 5 The actual roll angle of the drone is less than or equal to the given limit. The correctness of the proposed algorithm is demonstrated. In addition, the results obtained by the interior point method in the Matlab optimization toolbox are consistent with the results given by the optimization algorithm, proving the correctness of the proposed algorithm.

[0118] Figure 6 、 Figure 7 、 Figure 8 and Figure 9 The energy consumption comparison of the optimization method and the minimum and maximum flight speed of the UAV under different Q values, the actual amount of data received by the destination node, the flight speed of the UAV and the actual roll angle of the UAV are given respectively. Here, assuming L = 2000,

[0119] Figure 6 It is shown that the proposed optimization algorithm has performance advantages in terms of power consumption; Figure 7 It shows that the actual amount of data received by the destination node is greater than the set Q value, which shows the correctness of the proposed algorithm; Figure 8 It shows that the actual flight speed of the UAV meets the given constraints, which shows the correctness of the proposed algorithm; Figure 9 The actual roll angle of the drone is less than or equal to the given limit. The correctness of the proposed algorithm is demonstrated. In addition, the results obtained by the interior point method in the Matlab optimization toolbox are consistent with the results given by the optimization algorithm, proving the correctness of the proposed algorithm.

[0120] Figure 10 、 Figure 11 、 Figure 12 and Figure 13 Different The optimization method is used to compare the energy consumption at the minimum and maximum flight speeds of the UAV, the actual amount of data received by the destination node, the flight speed of the UAV, and the actual roll angle of the UAV. Here, it is assumed that L = 2000 and Q = 100.

[0121] Figure 10 It is shown that the proposed optimization algorithm has performance advantages in terms of power consumption; Figure 11 It shows that the actual amount of data received by the destination node is greater than the set Q value, which shows the correctness of the proposed algorithm; Figure 12 It shows that the actual flight speed of the UAV meets the given constraints, which shows the correctness of the proposed algorithm; Figure 13 The actual roll angle of the drone is less than or equal to the given limit. This demonstrates the correctness of the proposed algorithm. In addition, the results obtained by the interior point method in the Matlab optimization toolbox are consistent with the results given by the optimization algorithm, further proving the correctness of the proposed algorithm.

[0122] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for optimizing a full-duplex fixed-wing UAV relay system with roll angle limitation, characterized in that: The method comprises the following steps: Step 1: Within the flight radius given by equation (24b), use the one-dimensional search algorithm to solve problem (24a) and obtain the UAV flight radius r2; Among them, Q is the amount of data sent from the source node S to the destination node D, r is the radius of the UAV flying in a circle at a constant speed v, It is the roll angle limit of the UAV during flight. Less than or equal to 45°, V min is the minimum flight speed of the UAV, e represents the natural constant, σ 2 represents the variance of Gaussian white noise, is the power consumption of a fixed-wing UAV flying a circular trajectory at a constant speed v and radius r, where g represents the acceleration due to gravity and c1 = ηC D0 B / 2, c2=2W 2 / [(πe0A R )ηB], where η represents the air density, C D0 represents the zero lift drag coefficient, B represents the wing area, e0 is the wingspan efficiency, W represents the overall weight of the drone, A R represents the aspect ratio of the drone wing, Among them, P TX is the transmit power of S and R, and are the average channel gains from S to R and from R to D at time t, respectively; Step 2: Substitute r2 as r into equation (23) to obtain tanφ; Among them, φ is the actual roll angle of the UAV flight, and it is required Established; Step 3: If If true, calculate the corresponding flight speed at this time Otherwise jump to step 6; Step 4: If If it holds, then the optimal flight radius r * =r2, optimal flight speed v * It can be calculated by formula (25), and then jump to step 9; Among them, V max is the maximum flight speed of the UAV; Step 5: If The optimal flight speed v * =V min , optimal flight radius r * It can be calculated by formula (26), and then jump to step 9; Step 5: If If it holds, then the optimal flight speed v * =V max , optimal flight radius r * It can be calculated by formula (27), and then jump to step 9; Step 6: If Then, within the flight radius given by formula (28b), the one-dimensional search algorithm is used to solve problem (28a) to obtain the UAV flight radius r3; Step 7: If If it holds, then the optimal flight radius r * = r3, optimal flight speed v * It can be calculated by formula (29), and then jump to step 9; Step 8: If If true, the optimal flight speed is v * =V max , optimal flight radius r * It can be calculated by formula (30); Step 9: Set the optimal flight radius r * Substitute r into formula (21) to calculate the optimal flight time T * ; Where T is the time it takes for the drone to complete forwarding Q amount of data.

Images