Unmanned aerial vehicle emergency communication network coverage and energy efficiency optimization method

Abstract

The invention relates to the technical field of wireless communication, in particular to an unmanned aerial vehicle emergency communication network coverage and energy efficiency optimization method, which comprises the steps that an unmanned aerial vehicle network acquires a received signal strength indication signal; the user clustering and power distribution module generates user clustering information and power distribution information; the triangulation positioning module obtains a received signal strength indication signal based on the user clustering information, and outputs a user position estimation value; the unmanned aerial vehicle horizontal deployment module receives a user position estimation value based on the user clustering information, and outputs an unmanned aerial vehicle horizontal position, an unmanned aerial vehicle coverage radius and a user coverage rate; on one hand, the height optimization module receives the unmanned aerial vehicle horizontal position and the unmanned aerial vehicle coverage radius based on the power distribution information and outputs the unmanned aerial vehicle height position; and based on the unmanned aerial vehicle horizontal position and the unmanned aerial vehicle height position, an optimal unmanned aerial vehicle three-position deployment scheme is obtained. According to the method, low calculation overhead, high convergence speed and algorithm adaptability can be realized, and a better communication coverage effect is achieved.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technologies, and particularly to a method for optimizing the coverage and energy efficiency of an unmanned aerial vehicle (UAV) emergency communication network. Background Art

[0002] Natural disasters and accident disasters bring losses to the national economy and the safety of people's lives and property. In the complex post-disaster scenarios of "three disruptions" of power outage, road closure, and network disconnection, the public network becomes unavailable due to paralysis. Relying solely on portable communication devices or communication vehicles that are difficult to reach the hinterland of the disaster area to provide remote command communication services will result in extremely limited network coverage and communication support capabilities, thereby seriously affecting the rescue work. To improve the emergency communication guarantee ability and accelerate the construction of the emergency communication systemization and intelligence, it is urgent to propose a method for optimizing the coverage and energy efficiency of an unmanned aerial vehicle (UAV) network for emergency communication guarantee.

[0003] Currently, most of the UAV-based emergency communication network coverage methods adopt orthogonal multiple access technologies. When the locations of affected users are known, the advantages of UAVs that can break through the "three disruptions" limitations are used to optimize the deployment strategy. Among them, some solutions use mathematical methods for three-dimensional deployment of UAVs for emergency network coverage, which have a relatively fast algorithm convergence speed, but have problems such as poor scheme adaptability and high computational complexity. Other solutions model the optimization problem using a Markov decision process and use reinforcement learning to explore the optimal deployment decision, but have problems such as strong data dependence, large computational overhead, and long training time. Heuristic solutions have received extensive attention because they can achieve low-overhead, low-complexity, and highly adaptable UAV network deployments. However, in current emergency scenarios, existing heuristic UAV emergency communication network coverage solutions often regard the locations of affected users as known, which does not conform to the actual situation; at the same time, existing heuristic solutions usually have problems such as low user coverage rate, poor communication quality, and short battery life. This is because they only consider expanding the coverage area of the UAV network but ignore the contradiction between the number of covered users and the upper limit of the UAV service capacity, resulting in poor communication quality for users even though they are within the service range of the UAV; in addition, most existing solutions use traditional orthogonal multiple access technologies for communication, which cannot support the long-term battery life and low-energy consumption access of a large number of affected users and cannot guarantee the minimum communication requirements when the traffic surges.

[0004] To address the above problems, the present invention provides a method for optimizing the coverage and energy efficiency of an unmanned aerial vehicle (UAV) emergency communication network. Summary of the Invention

[0005] In view of the technical problems of the existing UAV emergency communication network coverage scheme, such as poor adaptability, high computational complexity, low user coverage rate, poor communication quality, and short battery life, a method for UAV emergency communication network coverage and energy efficiency optimization is provided.

[0006] The technical means adopted by the present invention are as follows: A method for UAV emergency communication network coverage and energy efficiency optimization, comprising the following steps: S1. The UAV broadcasts a positioning beacon signal to the user equipment, and the user equipment receives the positioning beacon signal broadcast and feeds back a received signal strength indication signal to the UAV network; S2. Construct a trained UAV network coverage and energy efficiency optimization model, which includes a user clustering and power allocation module based on non-orthogonal multiple access technology, a triangular positioning module based on received signal strength indication, a UAV horizontal deployment module based on an improved whale optimization algorithm, and a height optimization module based on an improved mirage algorithm; S3. On the one hand, the user clustering and power allocation module generates user clustering information and power allocation information based on the channel gains of the users covered by the UAV. The power allocation information includes inter-cluster power allocation information and intra-cluster power allocation information. On the other hand, it transmits the user clustering information to the triangular positioning module and the UAV horizontal deployment module, and transmits the power allocation information to the height optimization module; S4. The triangular positioning module obtains the received signal strength indication signal and outputs an estimated user position value; S5. The UAV horizontal deployment module receives the estimated user position value and outputs the UAV horizontal position, the UAV coverage radius, and the user coverage rate; S6. On the one hand, the height optimization module receives the UAV horizontal position and the UAV coverage radius and outputs the UAV height position. On the other hand, it transmits the UAV deployment information to the user clustering and power allocation module; S7. Based on the UAV horizontal position and the UAV height position, the optimal three-dimensional UAV deployment scheme is obtained.

[0007] Further, the user clustering and power allocation module generates user clustering information and power allocation information based on the channel gains of the users covered by the UAV, including: Arrange the channel gains of the users covered by the UAV in descending order; Adopt a clustering mechanism that maximizes the link gain difference to cluster the users covered by the UAV and generate user clustering information; Adopt the iterative water filling method to allocate power to each cluster and generate inter-cluster power allocation information; The power is allocated to the users in each cluster in a fixed - ratio distribution manner to generate the in - cluster power allocation information.

[0008] Furthermore, the calculation formula for the inter - cluster power allocation information is: , where, is the power allocated by the UAV to cluster h, is the water injection level, is the sum of the channel gains of the users in cluster h, is the Gaussian white noise power; The calculation formula for the in - cluster power allocation information is: , where, is the power allocation coefficient, is the power allocated by the UAV to cluster h, r is the ranking of the user in cluster h, is the efficiency of successive interference cancellation of the users in the user clustering and power allocation module.

[0009] Furthermore, the triangular positioning module obtains the received signal strength indication (RSSI) signal and outputs the user position estimate value, including: Inputting the RSSI signal value into the RSSI signal propagation model to calculate the distance between the user and the UAV; Selecting the three UAVs with the highest RSSI signal values and calculating the distance information between each of the three UAVs and the user; According to the distance information between each of the three UAVs and the user, constructing a non - linear equation and solving the non - linear equation by the least - squares method to obtain the user position estimate value.

[0010] Furthermore, the calculation formula for the distance between the user and the UAV is: ， where, is the reference distance, is the distance between the user and the UAV, n is the path loss exponent determined by the environment, is the RSSI signal value of the user, is the signal value at the reference distance ; The non - linear equation is: , where, is the estimated position of the user node, is the corresponding UAV position coordinate, is the distance between the UAV and the target user, .

[0011] Further, the S5 includes: S51. Calculate the coverage rate of the UAV network for ground users according to the number of users covered by the UAV and the total number of users; S52. Select the maximum number of iterations and the population size, and input the user position estimate value, UAV position, maximum number of iterations, and population size into the horizontally deployed communication coverage module; S53. Based on the UAV position, establish an initial population, and generate the positions of whale individuals according to the chaos mapping strategy; S54. Calculate the population fitness, and record the best fitness value and the corresponding position; S55. Calculate the non-linear convergence factor, vector A, vector C, set the random number and the probability random number set by the predation mechanism; S56. Judge whether the probability random number is less than 0.5. If the probability random number is less than 0.5, continue to judge whether the modulus of vector A is less than 1. If it is less than 1, execute S57. If it is greater than or equal to 1, execute S59; if the probability random number is greater than or equal to 0.5, execute S58; S57. Perform the behavior of surrounding and hunting prey, and execute S510 after completion; S58. Perform the bubble net predation behavior, and execute S510 after completion. The bubble net predation behavior improves the position update strategy by Levy flight; S59. Perform the random search for prey behavior, and execute S510 after completion; S510. Execute the elite selection behavior, generate the adaptive crossover probability, and generate new individuals through two-point crossover operation; S511. Adopt the elite retention strategy, directly retain the individual with the highest fitness in each generation to the next generation, select the individuals with the highest fitness to form a new population, calculate the fitness value of the population, and update the global optimal value; S512. Judge whether the maximum number of iterations is reached. If the maximum number of iterations is not reached, execute S55. If the maximum number of iterations is reached, execute S513; S513. Output the horizontal position of the UAV, the UAV coverage radius, and the user coverage rate.

[0012] Further, the calculation formula for the initial population is: ， where, is the original population, respectively represent the horizontal and vertical coordinates of the UAV and the coverage radius, represents the number of UAVs; The calculation formula for the position of the whale individual is: ， where, is the i th dimension position of the j th whale individual, is the value of the chaotic sequence in the j th dimension, is the upper bound of the j th dimension element, is the lower bound of the j th dimension element; The calculation formula for the position of the bubble net predation behavior is: , where, is a random number, is the current whale position, is the next position of the current whale, is vector A, is the distance between the current whale position and the current best whale position, is the Lévy flight random step size, is the dimension of the problem, is the distance between the current whale individual and the optimal individual; The calculation formula for the adaptive crossover probability is: ， where, is the adaptive crossover probability, is the initial crossover probability, is the final crossover probability, is the current iteration number, is the maximum iteration number; The generation method of the new individual is: ， where, is the new individual, is the position vector of the first parent individual, is the position vector of the second parent individual, is the randomly selected crossover point, is the dimension of the position vector, .

[0013] Furthermore, the S6 includes: S61. Determine the problem dimension, population size, maximum number of iterations, lower bound of the search space, and upper bound of the search space, and input the horizontal position of the UAV, the coverage radius of the UAV, the problem dimension, the population size, the maximum number of iterations, the lower bound of the search space, and the upper bound of the search space into the height optimization module; S62. Generate the initial population and positions according to the elite opposition-based learning strategy, calculate the fitness of the initial population, and record the best fitness value and the corresponding position; S63. Sort all the population individuals in descending order according to the fitness value; S64. Pass all the population individuals after the descending order through the upper mirage search strategy and the lower mirage search strategy in sequence, and output the lower mirage observation position; S65. Determine whether the current number of iterations has reached the maximum number of iterations. If not, jump to S64; S66. Output the set of UAV height positions, system energy efficiency, and the fitness value corresponding to the optimal solution.

[0014] Further, the S64 includes: S641. Calculate the number of individuals participating in the upper mirage search strategy, and update the individual position array p c according to the number of individuals. The calculation formula for the number of individuals is: , where is the number of individuals, is the maximum number of iterations, is the current number of iterations; S642. Calculate the vertical distance from the initial position to the horizontal refractive index manifold, the angle between the incident light of the initial position and the normal of its horizontal line, the angle between the refractive index stratified line and the ground, and determine the randomly selected direction coefficient. The calculation formula for the vertical distance from the initial position to the horizontal refractive index manifold is as follows: , where is the vertical distance of the th dimension of the th initial position, is the th dimension of the global optimal solution, is the solution of the th dimension of the th initial position, , The constraint range of is: is the maximum number of iterations, is the current number of iterations; The angle between the incident light at the initial position and the normal of the horizontal line where it is located and the angle between the refractive index stratification line and the ground The calculation formulas are as follows: , where, is the angle between the incident light at the initial position and the normal of the horizontal line where it is located, is the angle between the refractive index stratification line and the ground, ; S643. Calculate the superior mirage position increment according to the position relationship of the incident light; S644. Update the superior mirage observation position; S645. Determine whether all problem dimensions have been traversed. If not, jump to S642; if traversed, jump to S646; S646. Limit the candidate solution within the upper and lower bounds, and determine whether all individuals in the individual position array p c have been traversed. If not, jump to S641; if traversed, jump to S647 and output the global superior mirage optimal solution; S647. Select the optimal individual to update the current superior mirage population. The superior mirage search strategy is completed, and the inferior mirage search strategy is carried out; S648. Update the vertical distance from the initial position to the refractive index manifold horizontal according to whether the current individual is the global optimal individual of the inferior mirage population and update the corresponding direction of the vertical distance from the initial position to the refractive index manifold horizontal to obtain the updated vertical distance from the initial position to the refractive index manifold horizontal ; S649. Calculate the first angle, the second angle, the third angle, and the inferior mirage position increment. The calculation formula for the first angle is: ， where, is the first angle, is the maximum number of iterations, is the current number of iterations, , The calculation formula for the second angle is: ， where, is the second angle, is the first angle, , The calculation formula for the third angle is as follows: ， Wherein, is the third angle, is the second angle, is the first angle, is the refractive index magnitude of the medium below the refractive index stratification line, is the refractive index magnitude above the refractive index stratification line, and , The calculation formula for the inferior mirage position increment is: ， Wherein, is the inferior mirage position increment, is the third angle, is the second angle, is the first angle, is the vertical distance from the updated initial position to the refractive index manifold level; S6410. Update the inferior mirage observation position; S6411. Determine whether all individuals in the inferior mirage population have been traversed. If the traversal is not completed, jump to S648. If the traversal is completed, output the inferior mirage observation position.

[0015] Furthermore, calculating the upper layer position increment according to the positional relationship of the incident light includes: Calculating the superior mirage position increment according to the situation, where the situation includes the first situation, the second situation, and the third situation, The first situation is specifically: when the condition that the incident light is on the left side of the horizontal reference normal is satisfied, the superior mirage position increment is: ， Wherein, is the superior mirage position increment, is the randomly selected direction coefficient, is the angle between the incident light at the initial position and the normal of its horizontal line, is the angle between the refractive index stratification line and the ground, is the vertical distance from the initial position to the refractive index manifold level; The second situation is specifically: when the conditions that the incident light is on the right side of the horizontal reference normal, the angle between the refractive index stratification line and the ground is less than the angle between the incident light at the initial position and the normal of its horizontal line, and the angle between the incident light at the initial position and the normal of its horizontal line is less than are satisfied, the superior mirage position increment is: ， Among them, the upper mirage position increment, is the randomly selected direction coefficient, is the angle between the incident light at the initial position and the normal of the horizontal line where it is located, is the angle between the refractive index stratified line and the ground, is the vertical distance from the initial position to the horizontal of the refractive index manifold; The specific third case is: when it satisfies that the incident light is on the right side of the normal of the horizontal reference line, the angle between the refractive index stratified line and the ground is less than the angle between the incident light at the initial position and the normal of the horizontal line where it is located, and the angle between the incident light at the initial position and the normal of the horizontal line where it is located is less than Under the condition of, the upper mirage position increment is: ， Among them, the upper mirage position increment, is the randomly selected direction coefficient, is the angle between the incident light at the initial position and the normal of the horizontal line where it is located, is the angle between the refractive index stratified line and the ground, is the vertical distance from the initial position to the horizontal of the refractive index manifold; Updating the vertical distance from the initial position to the horizontal of the refractive index manifold and the corresponding direction of updating the vertical distance from the initial position to the horizontal of the refractive index manifold, including: judging whether the current individual is the global optimal individual of the inferior mirage population according to the fourth case and the fifth case, The fourth case includes: If the current individual is not the optimal individual of the inferior mirage population: ， Among them, is the th dimension of the th initial position after update, is the jth dimension of the global optimal solution, is the corresponding direction of updating the vertical distance from the initial position to the horizontal of the refractive index manifold, ; The fifth case includes: If the current individual is the optimal individual of the inferior mirage population ， Among them, is the th dimension of the Dimensions Is the corresponding direction for updating the vertical distance from the initial position to the refractive index manifold level .

[0016] Compared with the prior art, the present invention has the following advantages: The present invention determines the position of the affected user through a triangulation module based on Received Signal Strength Indicator (RSSI). Secondly, an improved whale optimization algorithm (IWOA)-based UAV horizontal deployment module is adopted to obtain the optimal horizontal deployment position of the UAV and maximize the network user coverage rate. Then, a height optimization module based on the improved mirage search optimization (IMSO) is used to determine the deployment height of all UAVs and maximize the system energy efficiency. This solution overcomes the problems of poor applicability, high complexity, and large computational overhead existing in the existing solutions through the improved meta-heuristic algorithm, supports the large-scale access of affected users using NOMA technology, and meets the long endurance, high user coverage rate, and high energy efficiency requirements of the UAV network in the emergency scenario through the collaborative deployment of multiple UAVs.

[0017] For the above reasons, the present invention can be widely promoted in the fields of wireless communication and the like. Description of the Drawings

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 Is the overall framework diagram of a method for optimizing the coverage and energy efficiency of a UAV emergency communication network according to the present invention.

[0020] Figure 2 Is the flowchart of the UAV horizontal deployment module based on IWOA according to the present invention.

[0021] Figure 3 Is the flowchart of the height optimization module based on IMSO according to the present invention. Detailed Embodiments

[0022] To enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units need not be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0024] The present invention provides an energy-efficient unmanned aerial vehicle (UAV) network communication coverage method that can support online positioning of affected users and large-scale user access in complex emergency scenarios. By decoupling the UAV network coverage and energy efficiency optimization problems, and respectively using two improved metaheuristic algorithms for solution. The present invention mainly uses a triangulation positioning module based on RSSI to determine the user location, combines an improved whale optimization algorithm (IWOA) to optimize the horizontal deployment, and uses an improved mirage algorithm (IMSO) to optimize the altitude and energy efficiency, so as to achieve the technical effects of improving the network coverage rate, enhancing the communication quality, extending the endurance time, and meeting the low-energy consumption access requirements of large-scale users.

[0025] To solve the problems of low user coverage rate and poor algorithm performance in the horizontal deployment of UAVs in the emergency communication network for communication coverage, the present invention solves the target optimization sub-problem by using a whale optimization algorithm improved by strategies such as inferior value replacement, chaotic mapping perturbation, and adaptive crossover operator, and designs a UAV horizontal deployment and communication coverage module based on IWOA.

[0026] In order to solve the problems of poor user communication quality and high system energy consumption in the height optimization of drones in emergency communication networks for communication coverage, the present invention designs a user clustering and power allocation module based on NOMA, and designs the allocation mechanism of drone radio spectrum and power resources in detail to optimize the communication energy consumption of the system. Then, the mirage algorithm improved by the elite reverse learning strategy is adopted to solve the target optimization sub-problem, and a height and energy efficiency optimization module based on IMSO is designed.

[0027] like Figures 1-3 As shown, the present invention provides a method for optimizing the coverage and energy efficiency of a drone emergency communication network, comprising the following steps: S1. The drone transmits a positioning beacon signal broadcast to the user device, the user device receives the positioning beacon signal broadcast and feeds back a received signal strength indication signal to the drone network.

[0028] S2. Construct a UAV network coverage and energy efficiency optimization model. The UAV network coverage and energy efficiency optimization model includes a user clustering and power allocation module based on non-orthogonal multiple access technology, a triangulation positioning module based on received signal strength indication, a UAV horizontal deployment module based on an improved whale optimization algorithm, and a height optimization module based on an improved mirage algorithm.

[0029] S3. The user clustering and power allocation module generates user clustering information and power allocation information based on the channel gain of the users covered by the drone. The power allocation information includes inter-cluster power allocation information and intra-cluster power allocation information. On the other hand, it transmits the user clustering information to the triangulation positioning module and the drone horizontal deployment module, and transmits the power allocation information to the height optimization module.

[0030] To solve the problem of high energy consumption when users access on a large scale, the present invention designs a NOMA-based user clustering and power allocation module to achieve inter-cluster and intra-cluster power control of drone power, effectively reducing communication energy consumption; the drone emergency communication system follows the NOMA protocol, optimizes drone power allocation by changing the communication access method, and achieves multiple users in the same frequency band through power domain multiplexing. Covered user collection Divided into multiple clusters , each user in the cluster shares the same frequency band. The present invention considers using the maximum link gain difference user clustering algorithm to cluster the UAV The covered users are dynamically clustered, an iterative water injection scheme is used to allocate power between clusters of drones, and a fixed ratio allocation method is used to allocate power to users within a cluster, thereby achieving higher system performance with lower algorithm complexity. The specific design steps of the user clustering and power allocation module based on NOMA are as follows: First, perform user sorting. Arrange the channel gains of the users covered by the UAV in descending order. Assume that the UAV can be divided into multiple clusters, and the users within each cluster adopt the classic two-user mode. The attributive of the channel gain is the communication channel between the UAV and the users it covers. That is to say, the channel gain here refers to the channel gain of the communication channel between the UAV and the users it covers.

[0031] Secondly, execute the user clustering mechanism that maximizes the link gain difference. Adopt the clustering mechanism that maximizes the link gain difference to cluster the users covered by the UAV and generate user clustering information. Specifically, the user with the largest channel gain (the first one) and the user with the smallest channel gain (the last one) are grouped into one cluster. Then, the second-best user (the second one) and the second-to-last user are grouped into one cluster, and so on. When there is a single remaining user, this remaining user enjoys all the power of the UAV.

[0032] Subsequently, execute the inter-cluster power allocation mechanism for the UAV. Use the iterative water-filling method to perform adaptive power allocation for each cluster, that is, allocate higher power to the clusters with better channel conditions and less power to the clusters with worse channel conditions, and generate inter-cluster power allocation information. The calculation formula for the inter-cluster power allocation information is: , where, is the power allocated by the UAV to cluster h, is the water-filling level, is the sum of the channel gains of the users within cluster h, is the Gaussian white noise power.

[0033] Finally, execute the intra-cluster power allocation mechanism for the UAV. Adopt the method of fixed ratio allocation to allocate power to the users within each cluster. Users with larger channel gains obtain a lower proportion of power, while users with smaller channel gains obtain a larger proportion of power. For the users within cluster h, the detailed power allocation is carried out in inverse proportion to the channel gain to generate intra-cluster power allocation information. The calculation formula for the intra-cluster power allocation information is: , where, is the power allocation coefficient, is the power allocated by the UAV to cluster h, r is the ranking of the user within cluster h (in descending order of channel gain), is the efficiency of serial interference cancellation for the users in the user clustering and power allocation module.

[0034] S4. The triangulation module obtains the received signal strength indication signal and outputs the estimated user position.

[0035] In actual situations, the precise location of user equipment is protected by privacy rights. Therefore, network operators cannot obtain the location information of users without the authorization of the users. Moreover, due to the sudden occurrence of disasters, the location information of users is unknown. Therefore, a triangulation positioning module based on RSSI is designed. A temporary communication and positioning network is formed by multiple drones. Each drone broadcasts a positioning beacon signal. Subsequently, the trapped disaster-affected user equipment automatically scans all such received signals and provides feedback. The drone ad-hoc network selects the three UAVs with the optimal RSSI signal values as the signal sources for triangulation positioning, and then locates the location information of the user through geometric methods. The specific design steps are as follows: First, according to the triangulation positioning principle based on RSSI, the RSSI signal propagation model is derived: , where is the reference distance is the path loss, n is the path loss exponent determined by the environment, represents the influence of uncertain factors such as environmental noise and multipath effects. The second term in the formula describes the signal attenuation caused by the increase in distance.

[0036] Secondly, the input received signal strength indication signal value is input into the received signal strength indication signal propagation model to calculate the distance between the user and the drone. The calculation formula for the distance between the user and the drone is: ， where is the reference distance, is the distance between the user and the drone, n is the path loss exponent determined by the environment, is the received signal strength indication signal value of the user, is the signal value at the reference distance .

[0037] Then, all drones broadcast positioning beacon signals, select the three drones (UAVs) with the highest received signal strength indication signal values, and calculate the distance information between the three drones and the user respectively.

[0038] Finally, according to the distance information between the three drones and the user respectively, a non-linear equation is constructed, and the non-linear equation is solved by the least squares method to obtain the estimated value of the user location. The non-linear equation is: , where is the estimated location of the user node, is the corresponding drone position coordinate, is the distance between the drone and the target user, .

[0039] S5. The drone horizontal deployment module receives the user position estimation value and outputs the drone horizontal position, drone coverage radius and user coverage rate.

[0040] Regarding the sub-problem of horizontal deployment of drones, since the existing solutions have the problems of low user coverage and poor optimization effect, the present invention adopts the UAV horizontal deployment and communication coverage module based on IWOA for optimization. In order to solve the problems of convergence to local optimum, poor global exploration ability, premature convergence, etc. exhibited by the existing classic whale optimization method in the face of complex multi-constraint optimization problems of horizontal deployment of drones, the present invention adopts strategies such as inferior value replacement, chaotic mapping perturbation, and adaptive crossover operator to improve the algorithm. The IWOA algorithm is then used to optimize the horizontal deployment strategy and user coverage of drones. The input of the algorithm is the estimated value set of user locations, the number of drones, and the maximum number of iterations and population size of the algorithm. The output is the horizontal position set of drones, the coverage radius of drones, and the user coverage of the drone network. The flow chart of the UAV horizontal deployment and user coverage optimization module based on IWOA is shown below. Figure 2 shown.

[0041] S5 is as follows: S51. Calculate the coverage rate of the drone network to ground users based on the number of users covered by the drone and the total number of users, and define the coverage rate of the drone network to ground users as follows: ， in, For drones Total number of terrestrial users covered, N is the total number of ground users.

[0042] S52. Select the maximum number of iterations and the population size, and input the user position estimate, the drone position, the maximum number of iterations and the population size into the horizontally deployed communication coverage module.

[0043] S53. Population initialization. Based on the position of the drone, an initial population is established, and the individual positions of whales are generated according to the chaotic mapping strategy. The improvement of population initialization using chaotic mapping is as follows: The calculation formula for the initial population is: ， in, For the original population, Respectively represent the horizontal and vertical coordinates of the drone and the coverage radius, Indicates the number of drones.

[0044] The calculation formula for the position of an individual whale is: ， where, is the i -th dimension position of the j -th individual whale, is the value of the chaotic sequence in the j -th dimension, is the upper bound of the j -th dimension element, is the lower bound of the j -th dimension element, and there is , where the dimension .

[0045] S54. Calculate the fitness of the population, and record the best fitness value and the corresponding position.

[0046] S55. Calculate the non-linear convergence factor , vector A, vector C, set the random number l and the probability random number set by the predation mechanism. The specific forms of vector A and vector C are as follows: ， ， ， where, is vector A, is vector C, and are random numbers in [0, 1], is the current iteration number, is the maximum iteration number.

[0047] S56. Judge whether the probability random number is less than 0.5. If the probability random number is less than 0.5, continue to judge whether the modulus of vector A is less than 1. If it is less than 1, execute S57. If it is greater than or equal to 1, execute S59; if the probability random number is greater than or equal to 0.5, execute S58.

[0048] S57. Perform the behavior of encircling prey. After completion, execute S510. Assume that the best whale individual is the prey, and other whales update their positions according to the position of the best whale individual. Then the behavior of encircling prey is as follows: ， ， Among them, is the current position of the whale, is the next position of the current whale, is vector A, is the distance between the current whale position and the current best whale position, is vector C.

[0049] S58. Perform bubble-net hunting behavior, and after completion, execute S510. There are two hunting behaviors during the process of the leading whale approaching the prey, namely encircling the prey and bubble-net attack. Therefore, the WOA algorithm will randomly generate a number according to the probability set by the hunting mechanism to select bubble-net hunting or shrinking the encirclement, and use Lévy flight to improve the position update strategy: , Among them, is a random number, , is the spiral shape constant, is the current position of the whale, is the next position of the current whale, is vector A, is the distance between the current whale position and the current best whale position, is the Lévy flight random step size, is the dimension of the problem, is the distance between the current whale individual and the optimal individual, .

[0050] S59. Perform random search for prey behavior, and after completion, execute S510. To ensure that all whales can fully search in the solution space, the algorithm updates the position according to the distance between whales to achieve the purpose of global search: ， ， Among them, represents the position of a randomly selected whale, is vector A, is the distance between the current whale position and the current best whale position, is vector C, is the next position of the current whale, is the position of the current whale.

[0051] S510. Execute the elite selection behavior. Simulate the crossover operation in the genetic algorithm through an adaptive crossover probability dynamic adjustment strategy: ， Among them, is the adaptive crossover probability, is the initial crossover probability, is the final crossover probability, is the current iteration number, is the maximum iteration number.

[0052] Generate the adaptive crossover probability, and through the two-point crossover operation, generate new individuals: ， Among them, is the new individual, is the position vector of the first parent individual, is the position vector of the second parent individual, is the randomly selected crossover point, is the dimension of the position vector, .

[0053] S511. Adopt the elitist retention strategy to directly retain the individual with the highest fitness in each generation to the next generation to avoid the loss of excellent genes. The optimal individual is merged with the new generation population, the individual with the highest fitness is selected to form a new population, the fitness value of the population is calculated, and the global optimal value is updated.

[0054] S512. Determine whether the maximum iteration number is reached. If the maximum iteration number is not reached, execute S55. If the maximum iteration number is reached, execute S513.

[0055] S513. Output the current global optimal solution of the population, that is, the horizontal position of the UAV, the coverage radius of the UAV, and the user coverage rate.

[0056] S6. The height optimization module on the one hand receives the horizontal position of the UAV and the coverage radius of the UAV and outputs the height position of the UAV; on the other hand, it transmits the UAV deployment information to the user clustering and power allocation module.

[0057] For the sub - problem of UAV altitude optimization, since the existing solutions have problems such as poor communication quality and short endurance time in the UAV emergency communication network, the present invention uses an IMSO - based UAV altitude and energy - efficiency optimization module for optimization. To solve the problems of poor global exploration ability and poor solution quality shown by the newly proposed mirage algorithm in the face of the complex multi - constraint optimization problem of UAV altitude optimization, an elite reverse learning strategy is used to improve the algorithm. Specifically, the elite direction learning strategy is used to improve the initialization of the population. Specifically, a reverse learning mechanism is introduced to generate reverse solutions, expanding the initial solution space of the algorithm. The introduction of reverse solutions is beneficial to expanding the diversity of the initial solutions of the population, but the generation of its solutions is blind and the effect is not the best. Therefore, an elite learning strategy is introduced for further improvement, and individuals with fitness values higher than the original solutions in the reverse solutions are retained. The present invention optimizes the UAV altitude and the overall system energy - efficiency based on NOMA through the IMSO algorithm. The inputs of the algorithm are the UAV horizontal position, coverage radius, problem dimension, population size, maximum number of iterations, lower bound of the search space, and upper bound, and the outputs are the optimal solutions (UAV altitude and energy - efficiency ratio) and their fitness values. The flow chart is as shown in Figure 3 shown below.

[0058] S6 is specifically as follows: S61. Define the system energy - efficiency as the ratio of the sum - rate of the UAV emergency communication network to the total system energy consumption. Determine the problem dimension, population size, maximum number of iterations, lower bound of the search space, and upper bound of the search space, and input the UAV horizontal position, UAV coverage radius, problem dimension, population size, maximum number of iterations, lower bound of the search space, and upper bound of the search space into the altitude optimization module.

[0059] S62. Generate the initial population and positions according to the elite reverse learning strategy.

[0060] S63. Calculate the fitness of the initial population, record the best fitness value and the corresponding position, and then sort all the population individuals in descending order according to the fitness value.

[0061] The initial solutions need to satisfy the constraints of each dimension. Therefore, the vector generation formula for a single individual is as follows: ， where, is the th dimension of the current individual, is the lower bound of the dimension, is the upper bound of the dimension, is a random number, and

[0062] Introduce a reverse learning mechanism to generate a reverse solution, where the reverse point corresponds to the initial solution Can be defined as . Extending it to the multi-dimensional initial solution space, the population Any individual The reverse solution of is formulated as follows: ， An elite learning strategy is introduced, and only the reverse solutions with fitness values greater than the original solutions are recognized as elites and retained, otherwise they will be abandoned, and then the improved population will be sorted.

[0063] S64. All individuals of the population arranged in descending order are subjected to the upper mirage search strategy and the lower mirage search strategy in turn.

[0064] S64 is specifically: S641. Calculate the number of individuals participating in the superior mirage search strategy , according to the number of individuals For individual position array p c To update, the formula for calculating the number of individuals is: , in, is the number of individuals, is the maximum number of iterations, is the current iteration number.

[0065] S642. Calculate the vertical distance from the initial position to the horizontal refractive index manifold , the angle between the incident light at the initial position and the normal of the horizontal line where it is located , the angle between the refractive index stratification line and the ground , and determine the randomly selected directional coefficient, the vertical distance of the initial position from the horizontal refractive index manifold The calculation formula is as follows: , in, For the The initial position of dimensions, The global optimal solution is dimensions, For the The initial position of The solution of the dimension, , The constraints are: , To ensure The maximum number of iterations that fluctuates within a non-zero and large range, is the current iteration number.

[0066] The angle between the incident light at the initial position and the normal of the horizontal line where it is located and the angle between the refractive index stratified line and the ground The calculation formulas are as follows: , where, is the angle between the incident light at the initial position and the normal of the horizontal line where it is located, is the angle between the refractive index stratified line and the ground, .

[0067] Set to indicate whether the current situation is on the left or right side of the normal of the horizontal reference line.

[0068] S643. Calculate the superior mirage position increment according to the position relationship of the incident light.

[0069] Specifically, calculate the angles A, B, C, D and the superior mirage position increment according to three different situations, including the first situation, the second situation and the third situation.

[0070] The first situation is specifically: when the condition that the incident light is on the left side of the horizontal reference line is met, the superior mirage position increment is: ， where, is the superior mirage position increment, is the randomly selected direction coefficient, is the angle between the incident light at the initial position and the normal of the horizontal line where it is located, is the angle between the refractive index stratified line and the ground, is the vertical distance from the initial position to the horizontal of the refractive index manifold.

[0071] The second situation is specifically: when the conditions that the incident light is on the right side of the horizontal reference line, the angle between the refractive index stratified line and the ground is less than the angle between the incident light at the initial position and the normal of the horizontal line where it is located, and the angle between the incident light at the initial position and the normal of the horizontal line where it is located is less than are met, the superior mirage position increment is: ， where, is the superior mirage position increment, is the randomly selected direction coefficient, is the angle between the incident light at the initial position and the normal of the horizontal line where it is located, is the angle between the refractive index stratified line and the ground, is the vertical distance from the initial position to the horizontal of the refractive index manifold.

[0072] The third case is specifically: when the incident light is on the right side of the horizontal reference normal, the angle between the refractive index stratification line and the ground is less than the angle between the incident light at the initial position and the normal of its horizontal line, and the angle between the incident light at the initial position and the normal of its horizontal line is less than Under the condition that, the upper mirage position increment is: ， where, the upper mirage position increment, is the randomly selected direction coefficient, is the angle between the incident light at the initial position and the normal of its horizontal line, is the angle between the refractive index stratification line and the ground, is the vertical distance from the initial position to the horizontal of the refractive index manifold.

[0073] S644. Update the upper mirage observation position: ， where, is the solution of the th dimension of the th initial position in the t-th iteration, is the solution of the th dimension of the th initial position in the (t + 1)-th iteration, the upper mirage position increment.

[0074] S645. Judge whether all problem dimensions have been traversed , if not, jump to S642, if completed, jump to S646.

[0075] S646. Limit the candidate solution within the upper and lower bounds, and judge whether all individuals in the individual position array p c have been traversed. If not, jump to S641, if completed, jump to S647 to output the global upper mirage optimal solution.

[0076] S647. Select the optimal individual to update the current upper mirage population. The upper mirage search strategy is completed, and the lower mirage search strategy is carried out.

[0077] S648. According to whether the current individual is the global optimal individual of the lower mirage population, update the vertical distance from the initial position to the horizontal of the refractive index manifold and update the corresponding direction of the vertical distance from the initial position to the horizontal of the refractive index manifold to obtain the updated vertical distance from the initial position to the horizontal of the refractive index manifold .

[0078] Specifically, according to the fourth case and the fifth case, it is determined whether the current individual is the globally optimal individual of the inferior mirage population. The fourth case includes: If the current individual is not the optimal individual of the inferior mirage population: ， wherein, is the th dimension of the updated th initial position, is the jth dimension of the global optimal solution, is the corresponding direction of the vertical distance from the updated initial position to the refractive index manifold level, , takes values of ±1. +1 represents the upper bound direction of the optimal solution individual in the current solution individuals, and -1 represents the lower bound direction of the optimal solution individual in the current solution individuals.

[0079] The fifth case includes: If the current individual is the optimal individual of the inferior mirage population , wherein, is the th dimension of the updated th initial position, is the corresponding direction of the vertical distance from the updated initial position to the refractive index manifold level, , takes values of ±1. +1 represents the upper bound direction of the optimal solution individual in the current solution individuals, and -1 represents the lower bound direction of the optimal solution individual in the current solution individuals.

[0080] S649. Calculate the first angle , the second angle , the third angle and the inferior mirage position increment . The calculation formula for the first angle is that the first angle is the angle between angle 0 and : ， wherein, is the first angle, is the maximum number of iterations, is the current number of iterations, .

[0081] The calculation formula for the second angle is: ， Among them, is the second angle, is the first angle, .

[0082] The calculation formula for the third angle is: ， Among them, is the third angle, is the second angle, is the first angle, is the refractive index magnitude of the medium below the refractive index stratification line, is the refractive index magnitude above the refractive index stratification line, and .

[0083] The calculation formula for the inferior mirage position increment is: ， Among them, is the inferior mirage position increment, is the third angle, is the second angle, is the first angle, is the vertical distance from the updated initial position to the horizontal of the refractive index manifold.

[0084] S6410. Perform the update of the inferior mirage observation position: ， Among them, is the inferior mirage position increment, is the t+1 th solution of the i th dimension of the j rd initial position in the th iteration, i is the j th solution of the

[0085] After that, the candidate solutions are restricted within the upper and lower bounds, and the global optimal solution is updated simultaneously.

[0086] S6411. Determine whether all individuals in the inferior mirage population have been traversed. If the traversal is not completed, jump to S648. If the traversal is completed, output the inferior mirage observation position.

[0087] S65. Determine whether the current iteration number has reached the maximum iteration number. If not, jump to S64.

[0088] S66. Output the current optimal solution of the population, where the optimal solution of the population includes the set of the altitude positions of the UAVs, the system energy efficiency, and the fitness value corresponding to the optimal solution.

[0089] S7. Based on the horizontal position and the altitude position of the UAVs, obtain the optimal three-dimensional deployment plan of the UAVs.

[0090] This method combines Non-Orthogonal Multiple Access (NOMA) technology for communication access. After locating the affected users, it decouples the problems of UAV network coverage and energy efficiency optimization, and respectively uses two improved meta-heuristic algorithms to solve the two sub-problems of UAV horizontal deployment and altitude optimization, so as to obtain the optimal three-dimensional deployment plan. The present invention determines the positions of the affected users through a triangulation module based on the Received Signal Strength Indicator (RSSI). Secondly, it uses a UAV horizontal deployment module based on the Improved Whale Optimization Algorithm (IWOA) to obtain the optimal horizontal deployment positions of the UAVs and maximize the network user coverage rate. Then, it uses an altitude optimization module based on the Improved Mirage Search Optimization (IMSO) to determine the deployment altitudes of all UAVs and maximize the system energy efficiency. This solution overcomes the problems of poor applicability, high complexity, and large computational overhead existing in the existing solutions through the improved meta-heuristic algorithms, supports the large-scale access of affected users using NOMA technology, and meets the requirements of long endurance, high user coverage rate, and high energy efficiency of the UAV network in the emergency scenario through the collaborative deployment of multiple UAVs.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing the coverage and energy efficiency of an unmanned aerial vehicle emergency communication network, characterized in that, It includes the following steps: S1. The drone transmits a positioning beacon signal broadcast to the user device, and the user device receives the positioning beacon signal broadcast and feeds back a received signal strength indication signal to the drone network; S2. Build a trained drone network coverage and energy efficiency optimization model, which includes a user clustering and power allocation module for non-orthogonal multiple access technology, a triangulation positioning module based on received signal strength indication, a drone horizontal deployment module based on an improved whale optimization algorithm, and an altitude optimization module based on an improved mirage algorithm; S3. On the one hand, the user clustering and power allocation module generates user clustering information and power allocation information based on the channel gains of the users covered by the drone. The power allocation information includes inter-cluster power allocation information and intra-cluster power allocation information. On the other hand, it transmits the user clustering information to the triangulation positioning module and the drone horizontal deployment module, and transmits the power allocation information to the altitude optimization module; S4. The triangulation positioning module obtains the received signal strength indication signal and outputs an estimated user position value; S5. The drone horizontal deployment module receives the estimated user position value and outputs the horizontal position of the drone, the coverage radius of the drone, and the user coverage rate; S6. On the one hand, the altitude optimization module receives the horizontal position of the drone and the coverage radius of the drone and outputs the altitude position of the drone. On the other hand, it transmits the drone deployment information to the user clustering and power allocation module; S7. Based on the horizontal position and altitude position of the drone, obtain the optimal three-dimensional drone deployment plan.

2. The method for optimizing the coverage and energy efficiency of an unmanned aerial vehicle emergency communication network according to claim 1, characterized in that, The user clustering and power allocation module generates user clustering information and power allocation information based on the channel gains of the users covered by the drone, including: Arrange the channel gains of the users covered by the drone in descending order; Adopt a clustering mechanism that maximizes the link gain difference to cluster the users covered by the drone and generate user clustering information; Adopt the iterative water filling method to allocate power to each cluster and generate inter-cluster power allocation information; Adopt a fixed ratio allocation method to allocate power to the users within each cluster and generate intra-cluster power allocation information.

3. The method for optimizing the coverage and energy efficiency of the drone emergency communication network according to claim 2, characterized in that The calculation formula for the inter-cluster power allocation information is: ， Among them, is the power allocated to the drone for cluster h, is the water injection level, is the sum of the channel gains of the users within cluster h, is the Gaussian white noise power; The calculation formula for the intra-cluster power allocation information is: ， Among them, is the power allocation coefficient, is the power allocated by the UAV to cluster h, r is the ranking of the user within cluster h, is the efficiency of successive interference cancellation of the user in the user clustering and power allocation module.

4. The method for optimizing the coverage and energy efficiency of an unmanned aerial vehicle emergency communication network according to claim 1, wherein The triangulation positioning module obtains the received signal strength indication signal and outputs an estimated user position value, including: Input the received signal strength indication signal value into the received signal strength indication signal propagation model to calculate the distance between the user and the drone; Select the three drones with the highest received signal strength indication signal values and calculate the distance information between each of the three drones and the user; According to the distance information between each of the three drones and the user, construct a non-linear equation and solve the non-linear equation by the least squares method to obtain the estimated user position value.

5. The method for optimizing the coverage and energy efficiency of an unmanned aerial vehicle emergency communication network according to claim 4, wherein, The calculation formula for the distance between the user and the drone is: ， Among them, is the reference distance, is the distance between the user and the drone, and n is the path loss exponent determined by the environment, is the received signal strength indication signal value of the user, is at the reference distance the signal value below; The non-linear equation is: ， Among them, is the estimated position of the user node, is the corresponding UAV position coordinate, is the distance between the UAV and the target user, .

6. The method for optimizing the coverage and energy efficiency of the UAV emergency communication network according to claim 1, wherein, The S5 includes: S51. Calculate the coverage rate of the drone network for ground users according to the number of users covered by the drone and the total number of users; S52. Select the maximum number of iterations and the population size, and input the user location estimate, the UAV location, the maximum number of iterations, and the population size into the horizontally deployed communication coverage module; S53. Based on the UAV location, establish an initial population, and generate the whale individual positions according to the chaotic mapping strategy; S54. Calculate the population fitness, and record the best fitness value and the corresponding position; S55. Calculate the non-linear convergence factor, vector A, vector C, set the random number, and the probability random number set by the predation mechanism; S56. Determine whether the probability random number is less than 0.

5. If the probability random number is less than 0.5, continue to determine whether the modulus of vector A is less than 1. If it is less than 1, execute S57. If it is greater than or equal to 1, execute S59; if the probability random number is greater than or equal to 0.5, execute S58; S57. Perform the behavior of encircling the prey, and execute S510 after completion; S58. Perform the bubble-net predation behavior, and execute S510 after completion. The bubble-net predation behavior improves the position update strategy using Levy flight; S59. Perform the behavior of randomly searching for the prey, and execute S510 after completion; S510. Execute the elite selection behavior, generate the adaptive crossover probability, and generate new individuals through two-point crossover operation; S511. Adopt the elite retention strategy, directly retain the individual with the highest fitness in each generation to the next generation, select the individuals with the highest fitness to form a new population, calculate the fitness value of the population, and update the global optimal value; S512. Determine whether the maximum number of iterations is reached. If the maximum number of iterations is not reached, execute S55. If the maximum number of iterations is reached, execute S513; S513. Output the horizontal position of the UAV, the UAV coverage radius, and the user coverage rate.

7. The method for optimizing the coverage and energy efficiency of the UAV emergency communication network according to claim 6, characterized in that, The calculation formula for the initial population is: ， Among them, is the original population, respectively represent the horizontal and vertical coordinates and the coverage radius of the UAV, represents the number of UAVs; The calculation formula for the whale individual position is: ， Among them, is the i -th j -dimensional position of the -th whale individual, j is the value of the chaotic sequence at the -th j -dimensional element, is the upper bound of the j -th -dimensional element; The calculation formula for the position of the bubble-net predation behavior is: ， Among them, is a random number, is the current position of the whale, is the next position of the current whale, is vector A, is the distance between the current whale position and the current best whale position, is the Lévy flight random step size, is the dimension of the problem, is the distance between the current whale individual and the optimal individual; The calculation formula for the adaptive crossover probability is: ， Among them, is the adaptive crossover probability, is the initial crossover probability, is the final crossover probability, is the current iteration number, is the maximum iteration number; The generation method of new individuals is: ， Among them, is a new individual, is the position vector of the first parental individual, is the position vector of the second parental individual, is a randomly selected crossover point, is the dimension of the position vector, .

8. The method for optimizing the coverage and energy efficiency of the UAV emergency communication network according to claim 1, wherein, The S6 includes: S61. Determine the problem dimension, population size, maximum number of iterations, lower bound of the search space, upper bound of the search space, and input the horizontal position of the UAV, the UAV coverage radius, the problem dimension, the population size, the maximum number of iterations, the lower bound of the search space, and the upper bound of the search space into the highly optimized module; S62. Generate the initial population and positions according to the elite opposition-based learning strategy, calculate the initial population fitness, and record the best fitness value and the corresponding position; S63. Sort all the population individuals in descending order according to the fitness value; S64. Pass all the population individuals after the descending order through the superior mirage search strategy and the inferior mirage search strategy in turn, and output the inferior mirage observation position; S65. Determine whether the current number of iterations reaches the maximum number of iterations. If not, jump to S64; S66. Output the set of the UAV height position, the system energy efficiency, and the fitness value corresponding to the optimal solution.

9. The method for optimizing the coverage and energy efficiency of an unmanned aerial vehicle emergency communication network according to claim 8, wherein The S64 includes: S641. Calculate the number of individuals participating in the superior mirage search strategy, and update the individual position array according to the number of individuals. p c The calculation formula for the number of individuals is as follows: ， wherein, is the number of individuals, is the maximum number of iterations, is the current iteration number; S642. Calculate the vertical distance from the initial position to the horizontal of the refractive index manifold , the angle between the incident light at the initial position and the normal of the horizontal line where it is located , the angle between the refractive index stratification line and the ground , and determine the randomly selected direction coefficient. The formula for calculating the vertical distance from the initial position to the horizontal of the refractive index manifold is as follows:​ ， wherein, is the vertical distance of the -th dimension of the -th initial position, is the -th dimension of the global optimal solution, is the solution of the -th dimension of the -th initial position, , The constraint range of is: where is the maximum number of iterations, and is the current number of iterations; The included angle between the incident light at the initial position and the normal line of its horizontal line and the included angle between the refractive index stratification line and the ground are calculated as follows: ， Among them, is the angle between the incident light at the initial position and the normal of its horizontal line, is the angle between the refractive index stratification line and the ground, ; S643. Calculate the superior mirage position increment according to the position relationship of the incident light; S644. Update the superior mirage observation position; S645. Determine whether all problem dimensions have been traversed. If not, jump to S642; if traversal is completed, jump to S646; S646. Limit the candidate solutions within the upper and lower bounds, and determine whether to traverse the individual position array p c for all individuals in it. If the traversal is not completed, jump to S641. If the traversal is completed, jump to S647 and output the globally optimal solution of the upper mirage; S647. Select the optimal individual to update the current upper mirage population. After the upper mirage search strategy is completed, perform the lower mirage search strategy; S648. Update the vertical distance from the initial position to the refractive index manifold level according to whether the current individual is the global optimal individual of the inferior mirage population and the corresponding direction of updating the vertical distance from the initial position to the refractive index manifold level to obtain the updated vertical distance from the initial position to the refractive index manifold level ; S649. Calculate the first angle, the second angle, the third angle, and the lower mirage position increment. The calculation formula for the first angle is: ， Among them, is the first angle, is the maximum number of iterations, is the current number of iterations, , The calculation formula for the second angle is: ， Among them, is the second angle, is the first angle, , The calculation formula for the third angle is: ， Wherein, is the third angle, is the second angle, is the first angle, is the refractive index magnitude of the medium below the refractive index stratification line, is the refractive index magnitude above the refractive index stratification line, and , The calculation formula for the lower mirage position increment is: ， Among them, is the inferior mirage position increment, is the third angle, is the second angle, is the first angle, is the vertical distance from the updated initial position to the refractive index manifold level; S6410. Update the lower mirage observation position; S6411. Determine whether all individuals in the lower mirage population have been traversed. If not, jump to S648; if traversal is completed, output the lower mirage observation position.

10. The method for optimizing the coverage and energy efficiency of an unmanned aerial vehicle emergency communication network according to claim 9, characterized in that, The calculation of the upper position increment according to the positional relationship of the incident light includes: Calculate the upper mirage position increment according to the situation, where the situation includes the first situation, the second situation, and the third situation, The first situation is specifically: when the condition that the incident light is on the left side of the horizontal reference normal is met, the upper mirage position increment is: ， Among them, the upper mirage position increment, is the randomly selected direction coefficient, is the angle between the incident light at the initial position and the normal of the horizontal line where it is located, is the angle between the refractive index stratification line and the ground, is the vertical distance from the initial position to the horizontal of the refractive index manifold; The second case is specifically as follows: when the incident light is on the right side of the horizontal reference normal line, the included angle between the refractive index stratification line and the ground is less than the included angle between the incident light at the initial position and the normal line of its horizontal line, and the included angle between the incident light at the initial position and the normal line of its horizontal line is less than Under the condition that, the position increment of the superior mirage is: ， Among them, the upper mirage position increment, is the randomly selected direction coefficient, is the angle between the incident light at the initial position and the normal of the horizontal line where it is located, is the angle between the refractive index stratification line and the ground, is the vertical distance from the initial position to the horizontal of the refractive index manifold; The specific third case is as follows: when the incident light is on the right side of the horizontal reference normal, the included angle between the refractive index stratification line and the ground is less than the included angle between the incident light at the initial position and the normal of its horizontal line, and the included angle between the incident light at the initial position and the normal of its horizontal line is less than under the condition that, the upper mirage position increment is: ， Among them, the upper mirage position increment, is a randomly selected direction coefficient, is the angle between the incident light at the initial position and the normal of the horizontal line where it is located, is the angle between the refractive index stratification line and the ground, is the vertical distance from the initial position to the horizontal of the refractive index manifold; Updating the vertical distance from the initial position to the refractive index manifold level according to whether the current individual is the global optimal individual of the inferior mirage population And updating the corresponding direction of the vertical distance from the initial position to the refractive index manifold level, including: judging whether the current individual is the global optimal individual of the inferior mirage population according to the fourth case and the fifth case The fourth situation includes: If the current individual is not the optimal individual in the lower mirage population: ， Among them, is the th dimension of the updated th initial position, is the jth dimension of the global optimal solution, is the corresponding direction for updating the vertical distance from the initial position to the refractive index manifold level, ; The fifth situation includes: If the current individual is the optimal individual in the lower mirage population ， Among them, is the th dimension of the updated th initial position, is the corresponding direction of the vertical distance from the updated initial position to the refractive index manifold level, .

Images