CN115103409A

CN115103409A - Resource allocation method for multi-beam unmanned aerial vehicle cooperative communication

Info

Publication number: CN115103409A
Application number: CN202210739762.8A
Authority: CN
Inventors: 杨冬; 曾孝平; 简鑫; 陈超; 吴浪云
Original assignee: Chongqing University
Current assignee: Chongqing University
Priority date: 2022-06-28
Filing date: 2022-06-28
Publication date: 2022-09-23

Abstract

The application provides a resource allocation method for multi-beam unmanned aerial vehicle cooperative communication. The resource allocation method comprises the steps of building a system model based on the number of unmanned aerial vehicles, the number of beams owned by each unmanned aerial vehicle, the coverage area of each beam, the coverage area divided in a gridding mode, the number of users in each grid and the positions of unmanned aerial vehicle base stations and the users in the grids; establishing a resource optimization model of the multi-beam unmanned aerial vehicle cooperative communication network based on the system model; and decoupling the resource optimization model into inner layer optimization and outer layer optimization by using an alternate joint optimization solution algorithm, wherein the inner layer optimization is used for optimizing the incidence relation between the user and the beam, power distribution and the position of the unmanned aerial vehicles, and the outer layer adopts a one-by-one reduction method to optimize the number of the unmanned aerial vehicles, so that the number of the unmanned aerial vehicles is minimized, the total power consumption of a network is minimized and the load among the unmanned aerial vehicles is balanced on the premise of meeting the communication requirement of the user. The method and the device optimize association between the beams of the unmanned aerial vehicles and the user set, power distribution, positions and number of the unmanned aerial vehicles.

Description

Resource allocation method for multi-beam unmanned aerial vehicle cooperative communication

Technical Field

The application relates to the technical field of communication, in particular to a resource allocation method for multi-beam unmanned aerial vehicle cooperative communication.

Background

The wireless communication network with the unmanned aerial vehicle serving as an aerial base station can dynamically and selectively increase the capacity of the network by virtue of the characteristics of rapid deployment, flexible relocation, good line-of-sight propagation path and the like, and the wireless communication network is considered to be one of basic capabilities of a 5G/6G cellular network. Also, a massive MIMO (multiple Input multiple Output) technology, which is one of the key technologies of future networks, is considered as a key technology for expanding the capacity of a system and improving the spectrum utilization rate. The millimeter waves with a very high carrier frequency can greatly reduce the physical size of the antenna elements so that the physical space requirements are less stringent, which provides the possibility for loading the antenna array on the drone. The multi-beam array antenna is combined with the multiple unmanned aerial vehicles, the unmanned aerial vehicles can search for a good channel environment by means of flexible mobility, and can also form directional beams by means of beam forming to improve signal power gain and suppress interference, so that the unmanned aerial vehicles can improve on-demand communication service for ground users more accurately and more efficiently, and the unmanned aerial vehicles have strong practical application significance in the fields of emergency communication and the like.

In order to fully exert the advantages of cooperative auxiliary communication of the unmanned aerial vehicle after the introduction of the multi-beam array antenna, basic network planning and deployment problems such as user-beam association, network total power consumption, unmanned aerial vehicle quantity and position optimization and the like are mainly researched. The user association problem is critical to the performance of the communication system because, regardless of the technology used in the communication system, it is first determined whether the access point and the user are associated before data transmission is performed. Some documents have studied the user association problem from the viewpoint of maximizing the user communication rate. And some documents consider the user association problem from the point of load balancing. The unmanned aerial vehicle is limited in self characteristics and energy, and the time that the unmanned aerial vehicle can hover in the air can directly influence the value of practical application of the wireless network of the unmanned aerial vehicle, so that a reasonable power control scheme is very important for the wireless network of the unmanned aerial vehicle. Some documents solve the power distribution problem of non-convex optimization by converting the power distribution problem into convex optimization. Some documents use the lagrangian dual method for power allocation problems. Some documents use successive convex approximation to approximate the optimal solution by solving an approximately convex optimization problem that is not a convex problem. Optimization of drone position is one of the most widely focused issues in current drone-related research. The performance of the whole communication network is greatly influenced by the planning of the spatial position of the unmanned aerial vehicle base station; the simplest is that if the drone is deployed in a location that does not cover the user, then no communication service can be provided. Some documents present a drone position deployment solution that maximizes the coverage area. Some literature studies how an unmanned aerial vehicle can achieve the best position for maximum reliability when acting as a repeater, with total power loss, total outage and total error rate as reliability indicators. Some documents maximize the sum of the downlink rates by optimizing the location and available resources of full-duplex drones. There is relatively little research on optimizing the number of drones. The most common is an estimation method, i.e. estimating the number of drones needed based on the number of users and the user needs. However, the above studies only consider the problem of single association that one user can only be associated with one base station; the multi-association problem introduced during multi-beam unmanned aerial vehicle cooperative communication is not considered, namely, one user can be associated to a plurality of service base stations; system performance gains due to multi-association flexibility cannot be adequately modeled.

Disclosure of Invention

One objective of the present application is to provide a resource allocation method for multi-beam unmanned aerial vehicle cooperative communication, so as to jointly optimize association between beams of unmanned aerial vehicles and a user set, power allocation, and positions and numbers of unmanned aerial vehicles, thereby finding a minimum number of unmanned aerial vehicles capable of meeting communication requirements of users, minimizing total power consumption of a network, and balancing loads among unmanned aerial vehicles.

To achieve the above and other related objects, some aspects of the present application provide a resource allocation method for multi-beam drone cooperative communication. The resource allocation method comprises the steps of building a system model based on the number of unmanned aerial vehicles, the number of beams owned by each unmanned aerial vehicle, the coverage area of each beam, the coverage area divided in a gridding manner, the number of users in each grid and the user positions of each user in an unmanned aerial vehicle base station and the grid; establishing a resource optimization model of the multi-beam unmanned aerial vehicle cooperative communication network based on the system model; decoupling the resource optimization model into inner layer optimization and outer layer optimization by using an alternating joint optimization solution algorithm, wherein the inner layer optimization optimizes incidence relation between users and beams, power distribution and unmanned aerial vehicle positions, and the outer layer optimization optimizes the number of unmanned aerial vehicles by adopting a one-by-one reduction method, so that the number of unmanned aerial vehicles is minimized, total network power consumption is minimized, and loads among the unmanned aerial vehicles are balanced on the premise of meeting user communication requirements.

In one embodiment, the system model includes: m of said drones; each unmanned aerial vehicle is provided with N beams which can be freely adjusted in pointing direction; the user area on the ground is subjected to gridding division according to the coverage range of the wave beam, and users belonging to the same grid after division are a user set K; counting the number U of users in each grid; establishing a three-dimensional Cartesian coordinate system by taking the boundary of a rectangular area where a user is located as an x axis and a y axis and taking the direction of an unmanned aerial vehicle i as a z axis; representing unmanned aerial vehicle base station [ x ] using the three-dimensional Cartesian coordinate system _i ,y _i ,H]And the position coordinate [ x ] of user k in the grid _k ,y _k ,0]M, N is a positive integer, and H represents a coordinate value of the unmanned aerial vehicle base station on the z axis.

In an embodiment, the resource optimization model is a nested multi-objective mixed integer non-convex optimization model, and includes the following formula:

the method comprises the following steps that (1) is a first objective function, the number of the unmanned aerial vehicles is required to be minimum, wherein M represents the number of the unmanned aerial vehicles, rho represents an association relation, p represents a power distribution relation, and [ x, y ] represents an unmanned aerial vehicle position relation;

equation (2) is a second objective function that requires the network total power consumption to be minimized; wherein p is _ij Represents the power allocated to the jth beam by the ith drone, N is the number of drone beams, and i and j are positive integers;

the formula (3) is a third objective function, and the load difference among all the unmanned aerial vehicle base stations is required to be minimum; where ρ is _ijk Representing the association between the kth grid point and the jth beam of the ith drone base station, S representing the number of sets of users,

is an intermediate variable, k is a positive integer;

formula (4) isA constraint indicating that the actual communication rate of each mesh must be greater than or equal to a communication rate requirement threshold, wherein,

wherein, g _ik Representing the channel gain between the ith unmanned plane and the kth user set, beta representing the channel power gain value under a specified reference distance of 1m, alpha representing the road loss coefficient, d _ik Denoted is the distance, δ, between drone i and user set j ² Is Gaussian white noise, R _th Minimum communication rate requirement for a single user, u _k Number of users, x, representing the kth set of users _i Is the coordinate value of the unmanned aerial vehicle i on the x axis, y _i Coordinate values of the unmanned aerial vehicle i on the y axis;

equation (5) is a second constraint, which means that each beam can only cover one grid;

the expression (6) shows that after the user set is covered by a plurality of unmanned aerial vehicle beams, the load is equally distributed to a plurality of beams covering the user set;

equation (7) is a third constraint, which is a power constraint and indicates that the maximum sum of the powers that each drone can allocate to the respective beams is p _max ；

Equation (8) is a fourth constraint that indicates that the associated variable can only be 0 or 1.

In an embodiment, the decoupling the resource optimization model into an inner-layer optimization and an outer-layer optimization by using an alternating joint optimization solution algorithm, where the inner-layer optimization is used to optimize the incidence relation between the user and the beam, the power allocation, and the position of the drone, and the outer-layer optimization optimizes the number of the drones by using a one-by-one reduction method, including:

decoupling the nested multi-objective mixed integer non-convex optimization model into the inner layer optimization and the outer layer optimization by using the alternative joint optimization solving algorithm;

the inner layer optimization is used for solving the optimal value of the rest group of variables through iteration after fixing two groups of incidence relation rho among the beams, power p distributed to one beam by the unmanned aerial vehicle and an unmanned aerial vehicle position coordinate [ x, y, H ], solving the user incidence relation, and then alternately optimizing power distribution and position deployment, and continuously reducing the total power consumption of the network until convergence; and

and the outer layer optimization reduces the number of the unmanned aerial vehicles one by one based on the result of the inner layer optimization until the minimum number of the unmanned aerial vehicles meeting the communication requirements of users is found.

In an embodiment, the step of solving the user association includes:

establishing a user association model; wherein the user association model is:

wherein, the first and the second end of the pipe are connected with each other,

and

solving the user association model by using an improved whale optimization algorithm to obtain an association relation between the beam and a user set, wherein f (rho) and h _k (p) belonging to an intermediate variable, η _k Is a penalty factor.

In one embodiment, the solving of the power allocation comprises:

establishing a power distribution model; wherein the power distribution model is:

the problem (13) is an upper bound of the original power allocation problem, wherein,

and

solving the power distribution model by using an interior point method to obtain the optimal power distribution, wherein h ^ub (p _ij ) Belonging to an intermediate variable, p _max Is the maximum transmission power of the unmanned aerial vehicle,

and distributing the power value of each beam for each unmanned aerial vehicle after l iterations.

In one embodiment, the solving of the location deployment comprises:

establishing a position deployment model; wherein the location deployment model is:

wherein the content of the first and second substances,

and

solving the position deployment model by using an improved differential evolution algorithm to obtain the optimal position distribution of the unmanned aerial vehicle, wherein the function

g _k (x, y) are intermediate variables,

as a penalty function, v is a penalty factor, ω _ij Are weight coefficients.

In an embodiment, the step of solving the user association model by using an improved whale optimization algorithm to obtain the association relationship between the drone beam and the user set includes:

step A-11, initializing whale position information, and mapping the whale position information into an association relation between the beam and the user set;

step A-12, calculating objective function values corresponding to whales at the moment and selecting the best whale as a whale head;

step A-13, simulating whale hunting rules to update the position information of each whale;

step A-14, the new position information is mapped into the associated variables again, and then objective function values corresponding to the whales are calculated;

step A-15, adjusting the whale population quantity according to a population scale self-adaptive strategy; and

and step A-16, repeating the steps A-13 to A-15 until the iteration is finished, and outputting the association relation between the beam and the user set.

In an embodiment, the step of solving the location deployment model by using an improved differential evolution algorithm to obtain an optimal location distribution of the drones includes:

b-11, initializing population information by using a K-means algorithm according to the position information of the user;

b-12, performing population crossing and variation according to a differential evolution algorithm rule;

b-13, calculating objective function values corresponding to all individuals of the population, and selecting the individuals by adopting a greedy strategy to obtain a new population;

and B-14, repeating the steps B-12 to B-13 until the iteration is finished, and outputting the position information corresponding to the individual with the minimum objective function value at the moment, namely the optimized position distribution of the unmanned aerial vehicle.

The alternating optimization algorithm comprises the following steps:

initializing the number of drones, the user location, and the power of each beam;

fixing power distribution and the unmanned aerial vehicle position, calculating the incidence relation between the wave beam and the user set according to an improved whale optimization algorithm, and calculating the network total power consumption of the unmanned aerial vehicle;

fixing the power distribution and the incidence relation, and optimizing the position of the unmanned aerial vehicle according to an improved differential evolution algorithm;

fixing the incidence relation and the unmanned aerial vehicle position, optimizing resource allocation according to an interior point method, calculating the total power consumption of the network and obtaining a reduction value of the total power consumption of the network;

if the drop value of the network total power consumption is larger than a preset threshold value, calculating the actual communication rate of each user set at the moment; and

outputting the number of drones, the association, the power allocation, and the drone location if the communication demand of each user is satisfied.

As described above, the resource allocation method for multi-beam drone cooperative communication according to one or more embodiments of the present application has the following beneficial effects:

in the multi-beam unmanned aerial vehicle cooperative communication network, in order to find the minimum number of unmanned aerial vehicles capable of meeting the communication requirements of users, minimize the total power consumption of the network and balance the load among the unmanned aerial vehicles, the association between the unmanned aerial vehicle beam and the user set, the power distribution, the positions and the number of the unmanned aerial vehicles are jointly optimized. And an alternate optimization scheme is provided to decouple the optimization problem into inner and outer layer optimization. The three sub-problems are optimized alternately by the inner layer, the beam user association problem of the unmanned aerial vehicles is solved by improving a whale algorithm, the power distribution problem is solved by a continuous convex approximation method, the position optimization problem is solved by improving a differential evolution algorithm, and finally the number of the unmanned aerial vehicles is reduced one by one on the alternate optimization result of the inner layer by the outer layer optimization to obtain the minimum number of the unmanned aerial vehicles. Numerical simulation shows that the proposed cooperation and joint optimization scheme of the multi-beam unmanned aerial vehicle can flexibly adapt to complex service demand distribution, reduce power consumption and balance loads among unmanned aerial vehicles.

Drawings

Figure 1 illustrates a model for multi-beam drone collaboration to provide on-demand communication services to ground users, in accordance with one embodiment of the present invention.

FIG. 2 is a block diagram illustrating a joint optimization scheme according to one embodiment of the invention.

Figure 3 illustrates drone load differences under the IBWA and max-RSS algorithms according to one embodiment of the present invention.

Fig. 4 is a graph illustrating the communication rate requirements versus actual communication rates for each set of users, in accordance with one embodiment of the present invention.

Fig. 5 shows power consumption of each drone under different user distributions according to one embodiment of the present invention.

Fig. 6 shows the sum of the actual communication rates of the users as a function of the number of iterations, according to an embodiment of the present invention.

Fig. 7 shows the variation of the sum of the actual communication rates of users with the number of iterations according to different algorithms of an embodiment of the present invention.

Fig. 8a is a diagram illustrating the number of positions of drones before and after joint optimization when users are evenly distributed according to an embodiment of the present invention.

Fig. 8b is a diagram illustrating the number of positions of the drones before and after the joint optimization when the users are in single gaussian distribution according to an embodiment of the present invention.

Fig. 8c is a diagram illustrating the number of positions of drones before and after the joint optimization in the multi-gaussian distribution of users according to an embodiment of the present invention.

Fig. 9 shows power consumption of drones distributed among different users according to one embodiment of the present invention.

Fig. 10 shows the total power consumption of the network as a function of the number of iterations, according to an embodiment of the invention.

Fig. 11 shows the total power consumption of a network as a function of the communication demand of the subscribers, according to an embodiment of the invention.

Fig. 12 shows the total power consumption of the network as a function of the number of users according to one embodiment of the invention.

Detailed Description

The following description of the embodiments of the present application is provided by way of specific examples, and other advantages and effects of the present application will be readily apparent to those skilled in the art from the disclosure herein. The present application is capable of other and different embodiments and its several details are capable of modifications and/or changes in various respects, all without departing from the spirit of the present application. It should be noted that the features in the following embodiments and examples may be combined with each other without conflict.

It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present application, and the drawings only show the components related to the present application and are not drawn according to the number, shape and size of the components in actual implementation, and the type, amount and proportion of each component in actual implementation may be changed arbitrarily, and the layout of the components may be more complicated. Moreover, in this document, relational terms such as "first," "second," and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.

In the multi-beam unmanned aerial vehicle cooperative communication network, in order to find the minimum number of unmanned aerial vehicles capable of meeting the communication requirements of users, minimize the total power consumption of the network and balance the load among the unmanned aerial vehicles, the association between the unmanned aerial vehicle beam and the user set, the power distribution, the positions and the number of the unmanned aerial vehicles are jointly optimized. And an alternate optimization scheme is provided to decouple the optimization problem into inner and outer layer optimization. The three sub-problems are optimized alternately by the inner layer, the beam user association problem of the unmanned aerial vehicle is solved by improving a whale algorithm, the power distribution problem is solved by a continuous convex approximation method, the position optimization problem is solved by improving a differential evolution algorithm, and finally the number of the unmanned aerial vehicles is reduced one by one on the inner layer alternate optimization result to obtain the required minimum number of the unmanned aerial vehicles by the outer layer optimization. Numerical simulation shows that the proposed cooperation and joint optimization scheme of the multi-beam unmanned aerial vehicle can flexibly adapt to complex service demand distribution, reduce power consumption and balance loads among unmanned aerial vehicles.

The application provides a resource allocation method for multi-beam unmanned aerial vehicle cooperative communication. The resource allocation method comprises the following steps:

step S110, based on the number of the unmanned aerial vehicles, the number of the beams owned by each unmanned aerial vehicle, the coverage area of each beam, the coverage area divided in a gridding mode, the number of users in each grid and the base station of the unmanned aerial vehicles and the user positions of each user in the grids, a system model is built.

And step S120, establishing a resource optimization model of the multi-beam unmanned aerial vehicle cooperative communication network based on the system model.

Step S130, decoupling the resource optimization model into inner layer optimization and outer layer optimization by using an alternating joint optimization solution algorithm, wherein the inner layer optimizes incidence relation between users and the beams, power distribution and unmanned aerial vehicle positions, and the outer layer optimization optimizes the number of the unmanned aerial vehicles by adopting a one-by-one reduction method, so that the purposes of minimizing the number of the unmanned aerial vehicles, minimizing total network power consumption and balancing loads among the unmanned aerial vehicles on the premise of meeting communication requirements of the users are achieved.

The resource allocation method for the multi-beam unmanned aerial vehicle cooperative communication will be described in detail in the following sections.

System model

Fig. 1 presents a network model of multi-beam drone collaboration to provide ad hoc communication services to ground users on demand. The figure contains M drones, I ═ I ₁ ,i ₂ ,i ₃ ,...,i _M Represents by "}; each unmanned aerial vehicle has N beams with pointing directions capable of being freely adjusted, and J is equal to { J ═ J ₁ ,j ₂ ,j ₃ ,...,j _N Represents it by "; the user area on the ground is divided in a gridding mode according to the coverage range of the unmanned aerial vehicle wave beam, the divided users belonging to the same grid are a user set, and K is used as { K ═ K ₁ ,k ₂ ,k ₃ ,...,k _s Represents it by "; the number of users included in each grid area is counted and recorded as U ═ U ₁ ,u ₂ ,u ₃ ,...,u _s }; establishing a three-dimensional Cartesian coordinate system by taking the boundary of a rectangular area where a user is located as an x axis and a y axis and taking the direction of an unmanned aerial vehicle i as a z axis, and respectively representing the position coordinates of an unmanned aerial vehicle base station and a grid point user as [ x ] and [ x ] respectively _i ,y _i ,H]And [ x ] _k ,y _k ,0]Wherein the grid point position is represented by the center point of each divided region.

Compared with the current system model taking a single user as the associated object, the system model after the grid division takes the user set in the grid as the associated object has the following advantages:

1) support for small-scale movement of users: as long as the user moving range does not exceed the grid to which the user belongs, the number of the users in the grid can not be changed, namely the weight value representing the number of the users can not be changed, so that the problems of readjustment of network deployment and consequent multiple calculations caused by frequent addition or quit of the users to network nodes due to the movement of the users are avoided.

2) Simplifying the beam pointing problem: each beam just covers one grid, so that the problem of beam pointing of the unmanned aerial vehicle can be converted into a correlation problem between the unmanned aerial vehicle beam and a user set for research, namely a user correlation problem;

3) simplifying the modeling mode of the association problem: when modeling the multi-association problem, the multi-association can be represented without applying the constraint of summing the association variables, namely, one user set can be covered by a plurality of beams;

4) simplified interference modeling: since a single beam can cover only one set of users, when considering interference power, interference from other signals can be ignored and only noise is considered, which simplifies formulation and problem solving.

In conclusion, the system model can support research on resource allocation problems of multi-beam unmanned aerial vehicle cooperative communication and has the advantage of simplifying problem modeling and solving.

Modeling assumptions

And then, gradually establishing a resource optimization model of the multi-beam unmanned aerial vehicle cooperative communication network on the basis of the system model. The modeling assumptions employed in the present application include: 1) the location distribution of the ground users is fixed and known; 2) the unmanned aerial vehicle has a fixed number of beams, and the adjustment of beam pointing is discretized, namely, the beams are directed from one grid to another grid, so that the overlapping condition cannot occur; 3) each beam of the drone can only cover one grid, but one grid can be covered by multiple beams simultaneously, i.e. multiple associations; 4) when the grid is covered by multiple beams, its load will be equally divided to the individual beams; 5) the number of users in each grid represents the current grid communication demand; 6) the path loss propagation model adopts a free space propagation model. Table 1 gives the main parameters used in the present application and their meanings.

TABLE 1 modeling notation and summary of its definitions

Nested multi-objective mixed integer non-convex optimization model

The method comprises the steps of establishing a nested multi-target mixed integer non-convex optimization model which comprehensively considers positions and number of unmanned aerial vehicles, power distribution and unmanned aerial vehicle beams and users, wherein the aims are to balance loads among the unmanned aerial vehicles, minimize total network power consumption and minimize the number of the unmanned aerial vehicles under the condition of meeting the communication requirements of the users, and the equations (1) to (8) are shown.

The meanings of the above formulae can be summarized as follows: 1) the formula (1) is a first objective function, and the number of unmanned aerial vehicles is required to be minimum; 2) equation (2) is a second objective function, requiring that the total power consumption of the network be minimized; 3) the formula (3) is a third objective function, and the load difference between the base stations of the unmanned aerial vehicles is required to be minimum (namely load balancing), wherein the load of the unmanned aerial vehicles is represented by the number of users related to the unmanned aerial vehicles; 4) equation (4) is a first constraint that indicates that the actual communication rate of each mesh must be greater than or equal to the communication rate requirement threshold, where each beam is given a unit bandwidth, where g _ik The calculation formula of (2) is as follows:

where β denotes a channel power gain value at a given reference distance of 1m, α denotes a path loss coefficient, d _ik The distance between the drone i and the user set k is represented; 5) equation (5) is a second constraint condition, which indicates that each beam can only cover one grid; 6) equation (6) indicates that after the user set is covered by a plurality of beams, the load is equally distributed to the plurality of beams covering the user set; 7) equation (7) is a third constraint, which is a power constraint and indicates that the maximum sum of the powers that each drone can allocate to the respective beams is p _max (ii) a 8) Equation (8) is a fourth constraint that indicates that the associated variable can only be 0 or 1. It is noted that no constraints are included herein

The meaning of this constraint is to limit each mesh to be covered by only one beam, and the absence of this constraint in the present model means that one mesh point is allowed to be covered by multiple beams, i.e. multi-association.

Alternative joint optimization solving algorithm

In the nested multi-target mixed integer non-convex optimization model, a plurality of continuous variables and integer variables are coupled together and are difficult to solve by adopting a standard convex optimization method, and the application provides a joint optimization solving algorithm based on alternative optimization, as shown in fig. 2. The nested optimization model is decoupled into an inner layer and an outer layer by the joint optimization solution algorithm, the inner layer jointly optimizes user association, power distribution and position deployment, and the outer layer optimizes the number of unmanned aerial vehicles. And the inner layer optimization is to fix two groups of the three groups of variables rho and p and the position [ x, y and H ] of the unmanned aerial vehicle, then iteratively solve the optimal values of the remaining group of variables, alternately optimize power distribution and position deployment after solving the user association relation, and continuously reduce the total power consumption of the network until convergence. And the outer layer optimization reduces the number of the unmanned aerial vehicles one by one based on the inner layer alternate optimization result until the minimum number of the unmanned aerial vehicles meeting the communication requirements of the users is found.

Sub-problem 1: optimal user association

A model and a solving method for a user association problem. After the positions, the number and the power of the fixed unmanned aerial vehicles are distributed, the problem of association between the beam-user sets is simplified by taking load balance among the unmanned aerial vehicles as a target and taking the requirement of user communication rate as a constraint:

the problem is a typical NP-hard problem, the meta-heuristic algorithm is very effective in solving the problem compared with the traditional algorithm, and a near-optimal solution can be found in polynomial time rather than exponential time. Before solving, firstly, the problem is transformed, and the constraint is converted into an objective function by adopting a punishment method, so that the problem is equivalent to:

wherein:

eta in the formula _k Is a penalty factor of 0 or more, when h _k (p) < 0, a large "penalty" is imposed on the augmented objective function so that the solution is not selected. Note of course that the penalty factor cannot be too large, otherwise the optimality of the resulting solution will be affected.

Whale Optimization Algorithm (WOA) is one of emerging meta-heuristic algorithms, and a near-optimal solution is solved by simulating a Whale bubble net hunting mode by utilizing a bionic principle. Because of this unique mechanism, WOA can provide good global search capabilities. The problem (12) is solved by improving a whale algorithm, the improved algorithm is named as IBWA (improved binding WOA for user association) algorithm, and the improvement points are mainly as follows:

1) constraints (5) and (8) belong to strong constraints, and a feasible solution is difficult to search by random search. Mapping the continuous whale position information into a related variable matrix rho according to a constraint (8) _MN,S ：

That is, the largest one of the values selected from each row of the matrix is 1, and the others are automatically set to 0, so that a binary associated variable matrix is obtained, which can be used to calculate the objective function value. Furthermore, according to the constraint (5), the present application also makes the variable with a value of 1 zero with a certain probability:

this probability will become progressively smaller from 0.5 as the number of iterations increases.

2) The whale population scale is changed in a self-adaptive mode, and the possibility of obtaining global optimum is increased by changing the size of a search space in a self-adaptive mode. The basic idea of the population adaptive strategy is as follows: and if the position information of the head whales is not updated in one iteration and the current population number is less than the set maximum population number, the number of the whales in the whale population is increased.

The IBWA algorithm flow is shown in table 2:

a-11, initializing whale position information, and mapping the whale position information into an incidence relation between an unmanned aerial vehicle beam and a user set;

step A-12, calculating objective function values corresponding to whales at the moment and selecting the best whale as a head whale;

step A-14, the new position information is mapped into the associated variables again, and then objective function values corresponding to whales at the moment are calculated;

step A-15, adjusting the whale population quantity according to a population scale self-adaptive strategy

And step A-16, repeating the steps A-13 to A-15 until the iteration is finished, and outputting the association relation between the unmanned aerial vehicle beam and the user set.

TABLE 2 IBWA Algorithm

Sub-problem 2: optimal power allocation

A model and a solution method for a power distribution problem. The position quantity of the fixed unmanned aerial vehicles and the incidence relation between the unmanned aerial vehicle wave beams and the user set are used for minimizing the total power consumption of the unmanned aerial vehicles as a target, and the problem of power distribution of the unmanned aerial vehicles which are constrained by the requirement of satisfying the user communication rate is simplified as follows:

due to the variable p in the first constraint _ij Is a concave function, so it is clear that the problem is not a convex optimization problem. The solution of the problem cannot be done directly using classical convex optimization theory. Order to

Definition of

Allocating power values of all beams to all unmanned planes after l iterations, and dividing h (p) _ij ) In that

The following inequality holds true for the taylor expansion:

thus, problems arise

Is an upper bound of the problem (21) in which

The problem (20) is typically a convex optimization problem because the objective function and the constraint are both associated variables p _ij Is a linear function of (a). For the convex optimization problem, a plurality of classical algorithms can be used for solving, and the method adopts an interior point method for solving. However, the problem (20) is an approximate transformation of the problem (17)And continuous convex approximation is also needed to approximately solve, and finally the local optimal solution of the original problem is obtained. The iterative algorithm for power allocation based on successive convex approximation is shown in table 3.

TABLE 3 Power Allocation Algorithm

Sub-problem 3: optimal location deployment

A model and solution method for a location deployment problem. Under the incidence relation among the fixed unmanned aerial vehicle quantity, the power distribution and the unmanned aerial vehicle wave beam and the user set, the optimal deployment problem of the unmanned aerial vehicle position is simplified as follows:

wherein the content of the first and second substances,

since the variables ρ and p are both determined, the feasible solution only needs to satisfy the constraint condition, and the problem can be transformed into:

wherein the content of the first and second substances,

v is a penalty factor, and a positive number which is large enough is taken. Here, some variation is made compared to the standard penalty function, such a transformation being related to the whole of the problem under consideration. Under the determination of the association relationship between beams and mesh points of the unmanned aerial vehicle and the bandwidth allocation of the beams, the requirement of the position optimization of the unmanned aerial vehicle for the sub-problem is only to meet the constraint (4), but more than one feasible solution meeting the requirement is met, and one of the final goals of combining the problems is to minimize the total power consumption of the network, so that a solution which can enable the communication rate of a user is found in all feasible solutions, and under the condition that the position of the unmanned aerial vehicle is as large as possible, the equivalent communication rate requirement of the user can be met with smaller power consumption, and further, the total power consumption can be reduced to the maximum extent during the joint optimization.

The differential evolution algorithm is a population-based adaptive global optimization algorithm, and has the advantages of few parameters and strong optimization capability compared with a genetic algorithm and an ant colony algorithm, and the optimal position deployment problem of the unmanned aerial vehicle is solved by adopting the differential evolution algorithm. Considering that initial information has a large influence on the possibility that the algorithm obtains the global optimal solution, the method improves a mechanism for randomly generating the initial information of the differential evolution algorithm, fully utilizes known information, namely user position distribution, and generates the initial unmanned aerial vehicle position information based on distance clustering by adopting a k-means algorithm. The improved differential evolution algorithm steps are shown in table 4:

Table 4 unmanned aerial vehicle position optimization algorithm

Iterative process of alternate joint optimization

Three sub-problem solving methods have been presented. The optimization method of the number of the unmanned aerial vehicles is that after the inner-layer joint optimization of user association, resource allocation and unmanned aerial vehicle position deployment is completed, the number of the unmanned aerial vehicles is reduced one by one until no matter how iterative optimization can not meet the user communication demand constraint under the given number of the unmanned aerial vehicles, and then the number of the unmanned aerial vehicles in the previous round is the minimum number of the unmanned aerial vehicles.

Table 5 gives the overall alternative joint optimization algorithm. Under the estimated number of the unmanned aerial vehicles, the incidence relation between the user set and the beams of the unmanned aerial vehicles is solved, and then power distribution and unmanned aerial vehicle position deployment are alternately optimized to continuously reduce the total power consumption of the network until convergence; and then reducing the number of the unmanned aerial vehicles one by one and performing the joint optimization of the three inner-layer sub-problems again. It should be noted that the alternative optimization algorithm is approximately optimal, and the optimal performance of the alternative optimization algorithm is affected by the setting of initial variables, such as the initial position of the drone.

TABLE 5 Alternatives optimization Algorithm

Numerical analysis

Numerical simulation results are mainly described and analyzed. The simulation scene considers a ground area, a plurality of unmanned planes with fixed beam number are suspended in the high altitude of 100m of the area, a downlink cellular network formed by the unmanned planes provides temporary communication service for users on the ground, and the users on the ground are divided into 9 user sets according to the coverage range of the beams. The number of drones is four without particular description, and each drone is fixed with three beams that can specify directions. Table 6 lists the simulation parameter settings common to each algorithm.

TABLE 6 simulation parameters

User association

The effectiveness of the proposed IBWA algorithm targeting load balancing with the user communication rate requirement as constraint is mainly verified from two aspects. In addition to the general parameters, other settings were: the number of the unmanned aerial vehicles is four, and the positions are randomly generated; the power distributed to each wave beam by the unmanned aerial vehicle is determined according to an average distribution mechanism; three conditions are designed for the distribution of users, namely uniform distribution, single Gaussian distribution and multi-Gaussian distribution; the communication rate requirement of a single user is 0.15 bit/s/Hz. And (5) carrying out simulation according to the algorithm 1 to obtain simulation data.

Fig. 3 illustrates the load difference between drones under different user distributions under the IBWA algorithm and the correlation mechanism based on the strongest received signal (max-RSS), where the dashed blue line is the ordinate scale corresponding to the right. From fig. 3, the following information can be obtained:

1) under the max-RSS association mechanism, the load difference among the unmanned aerial vehicles is continuously increased along with the increase of the number of users. Especially, when the users are distributed in a single-Gaussian mode and a multi-Gaussian mode, the load difference among the unmanned aerial vehicles can rapidly rise along with the increase of the number of the users, and the overload of individual base stations can easily cause the great reduction of the overall performance of the network; under the load balancing association mechanism provided by the application, even under the condition of single Gaussian distribution of users, the load difference between the unmanned aerial vehicles only slowly increases along with the increase of the number of users, and cannot rapidly increase along with the increase of the number of users;

2) compared with a max-RSS mechanism, the IBWA load balancing association mechanism provided by the application can obviously reduce the load difference among the unmanned aerial vehicles, and the advantages of the load balancing are more and more obvious along with the increase of the number of users.

Fig. 4 depicts the actual communication rate of each user set in different user distributions compared with the communication rate requirement thereof, wherein the solid line represents the actual communication rate of the user set, and the dotted line represents the communication rate requirement of the user set. As can be seen in fig. 4:

1) whether the users are subject to a uniform distribution, a single gaussian distribution or a multiple gaussian distribution, the solid lines in the figure are higher than the dashed lines, which means that the communication rate requirements of each set of users are satisfied;

2) the variation trend of the solid line representing the actual communication speed of the user set is basically consistent with the variation trend of the dotted line representing the communication speed demand of the user, which shows that the load balancing association algorithm provided by the application can determine the association relation according to the actual communication demand of the user, and further proves the effectiveness of the algorithm provided by the application;

3) the actual communication rate of some user sets is much greater than the required communication rate, while the actual communication rate of some other user sets is just slightly greater than the required communication rate, which indicates that there is some optimization space for power averaging allocation, and this is described in detail in the following of this application.

Power distribution

The effectiveness of the power allocation algorithm is verified. The simulation parameters are similar to the above, where the position of the drone and the association between the drone beam and the user are set according to the optimization results in the numerical analysis.

Fig. 5 depicts the total power consumption of each drone for different user distributions after the power optimization proposed by the present application, compared to the average power distribution, where the bar graph with the letter d represents the power average distribution. It can be seen from fig. 5 that the total power consumption of the unmanned aerial vehicle decreases regardless of the distribution to which the user obeys after power optimization, where the minimum decrease is the unmanned aerial vehicle 3 in multi-gaussian distribution, the total power consumption decreases by about 0.3W after optimization, and the maximum decrease is the unmanned aerial vehicle 1 in single-gaussian distribution, and the total power consumption decreases by about 0.6W.

Table 7 counts the total power consumption of the drone network after power optimization at different individual user communication rate requirements. From table 7, the following information can be obtained:

1) the speed requirement of a fixed user is compared with the total power consumption of the unmanned aerial vehicle under the distribution of different users, so that the power optimization effect is best when the users are in single Gaussian distribution, and the users are in Gaussian distribution for multiple times, but the total power consumption is lower than that when the users are in uniform distribution.

2) It can be found from fig. 4 that when the power is distributed averagely, the total power consumption of the network of 4W is the situation just meeting the communication rate requirement of a single user at 0.15bit/s/Hz, and after the power distribution algorithm provided by the present application is optimized, the total power consumption of the network required to meet the same communication rate requirement can be reduced to 2.21W, 2.09W and 2.12W, which fully explains the effectiveness of the power distribution optimization algorithm.

Table 7 total power consumption of network under different communication rate requirements

Location deployment

The performance of algorithm 3 is of great importance. Fig. 6 depicts the variation of the sum of the actual communication rates of the users with the variation of the number of iterations in different user distribution situations, where three solid lines consisting of circles, squares, and diamonds represent the user obeys the situations of uniform distribution, single gaussian distribution, and multiple gaussian distribution, respectively. As shown in fig. 6, it can be seen that:

1) with the increase of the iteration times, the sum of the actual communication rates of the users is also slowly increased, which shows the effectiveness of the algorithm, and a more appropriate unmanned aerial vehicle deployment position can be found so that the communication rate of the users is increased;

2) compare in user evenly distributed, when user single gauss distributes and many gauss distribute, objective function value increase range is bigger, and this shows that when user single gauss distributes and many gauss distribute, can obtain better effect through adjusting the unmanned aerial vehicle position.

The improved differential evolution algorithm is compared with the same type of particle swarm and genetic algorithm, and the realization of three solid lines consisting of circles, squares and diamonds in fig. 7 represents the improved differential evolution algorithm, the genetic algorithm and the particle swarm algorithm respectively. Compared with the genetic algorithm and the particle swarm algorithm, the improved differential evolution algorithm basically converges to the same value with the other two algorithms, but can achieve the convergence under the condition of fewer iterations.

Number of unmanned aerial vehicles

The effectiveness of unmanned aerial vehicle quantity optimization is mainly verified. The simulation scene setting is basically consistent with the content, when the optimal number of the unmanned aerial vehicles is searched, the initial number of the unmanned aerial vehicles is 5, and the positions are randomly distributed.

Fig. 8 depicts the situation of the distribution of the number and positions of drones before and after the joint optimization, wherein the dots located in the plane represent the users, and the solid circles and solid diamonds in the sky represent the positions of drones before and after the joint optimization, respectively. From fig. 8, the following information can be obtained:

1) after the number of unmanned aerial vehicles is optimized, when users are uniformly distributed, the requirement of user communication rate is met, the minimum number of unmanned aerial vehicles is 3, and the number of unmanned aerial vehicles is 4 under the condition of single Gaussian distribution and multi-Gaussian distribution of the users. This shows that under the same communication requirements, the distribution of users can affect the network service performance, and the users can meet the communication requirements with fewer unmanned aerial vehicles when uniformly distributed;

2) the optimal deployment position of the unmanned aerial vehicle is related to the user distribution, and the optimal deployment position of the unmanned aerial vehicle can be distributed from uniform distribution to concentrated distribution along with the change of the user distribution.

In addition, in order to further prove the correctness of the optimization of the number of the unmanned aerial vehicles, fig. 9 also counts the power consumption required by each unmanned aerial vehicle to meet the user communication rate requirement when the number of the unmanned aerial vehicles is 3. Wherein when the single gaussian distribution of user, the power consumption value of unmanned aerial vehicle 2 is too big, consequently reduces this value by 10 times.

It can be seen from the figure that when the number of the unmanned aerial vehicles is 3, the power consumption of each unmanned aerial vehicle is between 0.6W and 0.8W and is less than the maximum power value in order to meet the user requirements of uniform distribution; when the users are in multi-Gaussian distribution and single Gaussian distribution, the power consumption of the unmanned aerial vehicle exceeds the maximum value of 1W; especially in the case of a single gaussian distribution of users, the power required by the drone 2 is close to 10W, which is much greater than the maximum power value. Therefore, the requirement of the user communication rate under single-Gaussian distribution and multi-Gaussian distribution is met, and the number of the unmanned aerial vehicles is at least 4.

Network total power consumption

And verifying the effectiveness of the inner-layer alternation optimization algorithm in reducing the network power consumption. Fig. 10 illustrates the variation of the total power consumption of the network of drones with the number of iterations after the joint optimization of user association, power allocation and drone position under the optimal number of drones, wherein the number of drones is also set to 4 when evenly distributed for uniformity. As can be seen from the figure:

1) after joint optimization, the total power consumption of the network is continuously reduced along with the increase of the iteration times and finally converged. Under the conditions of uniform distribution, single Gaussian distribution and multi-Gaussian distribution of users, the total power consumption is respectively reduced to about 1.9W, 1.8W and 2W from the original 4W;

2) the joint optimization effect is related to the user distribution, wherein the effect is optimal when the user obeys a single gaussian distribution, and the effect is inferior but better when the user obeys a uniform distribution than when the user obeys a multi-gaussian distribution.

With reference to fig. 9, it is found by comparison that in the case that the number of the drones is 4, the total power consumption required is 1.89W and 2W, respectively, and the total power consumption of the network is greatly reduced. For further analysis, table 8 counts the number of beams associated with each user set under different user distributions. It can be seen that the beams and the user sets are multiple associations, and when a certain user set contains a large number of users, such as the user set 5 with a single gaussian distribution of users, the flexibility of multiple associations enables multiple beams to be used to cover the user set so as to meet the high communication requirement of the user set, thereby avoiding the high power consumption caused by using single beam to cover, and verifying the superiority of cooperative communication.

TABLE 8 number of beams associated with each user set

In order to further verify the necessity of joint optimization, the method compares the joint optimization with only power optimization, and compares the joint optimization with the power optimization from two aspects of different user communication speed requirements and different user numbers respectively. Fig. 11 depicts the change of the total power consumption of the unmanned aerial vehicle under different communication rate requirements, and fig. 12 counts the change of the total power consumption of the network after joint optimization under different users. The following conclusions can be summarized from fig. 11 and 12:

regardless of the distribution to which the users are subjected, the total power consumption of the network increases as the communication rate demand of a single user increases or as the number of users increases, but the solid line is always lower than the dotted line, which means that the total power consumption of the network can be reduced to a greater extent by joint optimization than by power optimization alone, with different numbers of users and different communication rate demands of users.

In multi-beam unmanned aerial vehicle cooperative communication network, in order to find the minimum number of unmanned aerial vehicles capable of meeting the communication requirements of users, minimize the total power consumption of the network and balance loads among the unmanned aerial vehicles, the application jointly optimizes the association between the unmanned aerial vehicle beam and the user set, power distribution, the positions and the number of the unmanned aerial vehicles. And an alternate optimization scheme is provided to decouple the optimization problem into inner and outer layer optimization. The three sub-problems of optimizing transaction in the inner layer are solved by improving a whale algorithm, the beam user association problem of the unmanned aerial vehicle is solved, the power distribution problem is solved by a continuous convex approximation method, the position optimization problem is solved by improving a differential evolution algorithm, and finally the number of the unmanned aerial vehicles is reduced one by one on the inner layer alternate optimization result to obtain the required minimum number of the unmanned aerial vehicles in the outer layer optimization. Numerical simulation shows that the proposed multi-beam unmanned aerial vehicle cooperation and joint optimization scheme can flexibly adapt to complex service demand distribution, reduce power consumption and balance loads among unmanned aerial vehicles.

The above embodiments are merely illustrative of the principles and utilities of the present application and are not intended to limit the application. Any person skilled in the art can modify or change the above-described embodiments without departing from the spirit and scope of the present application. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical concepts disclosed in the present application shall be covered by the claims of the present application.

Claims

1. A resource allocation method for multi-beam unmanned aerial vehicle cooperative communication is characterized by comprising the following steps:

building a system model based on the number of the unmanned aerial vehicles, the number of beams owned by each unmanned aerial vehicle, the coverage area of each beam, the coverage area divided in a gridding manner, the number of users in each grid and the base station of the unmanned aerial vehicles and the user positions of each user in the grids;

establishing a resource optimization model of the multi-beam unmanned aerial vehicle cooperative communication network based on the system model; and

decoupling the resource optimization model into inner layer optimization and outer layer optimization by using an alternative joint optimization solution algorithm, wherein the inner layer optimization is used for optimizing the incidence relation between users and the beams, power distribution and unmanned aerial vehicle positions, and the outer layer optimization adopts a one-by-one reduction method to optimize the number of unmanned aerial vehicles, so that the number of unmanned aerial vehicles, the total network power consumption and the load among the unmanned aerial vehicles are minimized on the premise of meeting the communication requirements of the users.

2. The method for resource allocation for multi-beam drone cooperative communication according to claim 1, wherein said system model comprises: m of said drones; each unmanned aerial vehicle is provided with N beams which can be freely adjusted in pointing direction; the user area on the ground is subjected to gridding division according to the coverage range of the wave beam, and users belonging to the same grid after division are a user set K; counting the number U of users in each grid; establishing a three-dimensional Cartesian coordinate system by taking the boundary of a rectangular area where a user is located as an x axis and a y axis and taking the direction of an unmanned aerial vehicle i as a z axis; representing unmanned aerial vehicle base station [ x ] using the three-dimensional Cartesian coordinate system _i ,y _i ,H]And the position coordinate [ x ] of user k in the grid _k ,y _k ,0]M, N is a positive integer, and H represents a coordinate value of the unmanned aerial vehicle base station on the z axis.

3. The method for resource allocation for multi-beam drone cooperative communication according to claim 2, wherein the resource optimization model is a nested multi-objective mixed integer non-convex optimization model comprising the following formula:

the method comprises the following steps that (1) is a first objective function, the number of unmanned aerial vehicles is required to be minimum, wherein M represents the number of the unmanned aerial vehicles, rho represents an association relation, p represents a power distribution relation, and [ x, y ] represents an unmanned aerial vehicle position relation;

equation (2) is a second objective function that requires the total power consumption of the network to be minimized; wherein p is _ij Represents the power allocated to the jth beam by the ith drone, N is the number of drone beams, and i and j are positive integers;

the formula (3) is a third objective function, and the load difference among all unmanned aerial vehicle base stations is required to be minimum; where ρ is _ijk Representing the association between the kth grid point and the jth beam of the ith drone base station, S representing the number of sets of users,

is an intermediate variable, k is a positive integer;

equation (4) is a first constraint that indicates that the actual communication rate for each mesh must be greater than or equal to the communication rate requirement threshold, wherein,

wherein, g _ik Representing the channel gain between the ith unmanned plane and the kth user set, beta representing the channel power gain value under a specified reference distance of 1m, alpha representing the road loss coefficient, d _ik Denoted is the distance, δ, between drone i and user set k ² Is Gaussian white noise, R _th Minimum communication rate requirement for a single user, u _k Number of users, x, representing the kth set of users _i Is the coordinate value of the unmanned aerial vehicle i on the x axis, y _i Coordinate values of the ith unmanned aerial vehicle on the y axis;

4. The method for resource allocation for multi-beam drone cooperative communication according to claim 3, wherein said decoupling said resource optimization model into an inner optimization and an outer optimization using an alternating joint optimization solution algorithm, said inner optimization for optimizing user association with said beam, power allocation and drone position, said outer optimization optimizing said number of drones using a one-by-one reduction method comprises:

the inner layer optimization is used for solving the optimal value of the rest group of variables through iteration after fixing two groups of incidence relation rho among the beams, power p distributed to each beam by the unmanned aerial vehicle and the position coordinate [ x, y, H ] of the unmanned aerial vehicle, solving the user incidence relation, then alternately optimizing power distribution and position deployment, and continuously reducing the total power consumption of the network until convergence; and

the outer layer optimization reduces the number of the unmanned aerial vehicles one by one based on the result of the inner layer optimization until the minimum number of the unmanned aerial vehicles meeting the communication requirements of users is found.

5. The method for resource allocation for multi-beam drone cooperative communication of claim 4, wherein the step of solving for the user association comprises:

establishing a user association model; wherein the user association model is:

wherein the content of the first and second substances,

and

6. The method for resource allocation for multi-beam drone cooperative communication according to claim 4, characterized in that said step of solving for power allocation comprises:

the problem (13) is an upper bound of the original power allocation problem, in which,

and

and distributing the power value of each wave beam for each unmanned aerial vehicle after l iterations.

7. The method for resource allocation for multi-beam drone cooperative communication according to claim 4, wherein said step of solving for a location deployment comprises:

wherein the content of the first and second substances,

and

g _k (x, y) are intermediate variables,

v is a penalty factor, ω _ij Are weight coefficients.

8. The method for resource allocation for multi-beam drone cooperative communication of claim 5, wherein said step of solving said user association model using a modified whale optimization algorithm to obtain an association between a drone beam and a set of users comprises:

step A-11, initializing whale position information, and mapping the whale position information into an incidence relation between the beam and the user set;

9. The method for resource allocation for multi-beam drone cooperative communication according to claim 7, wherein said step of solving said location deployment model using a modified differential evolution algorithm to obtain an optimal distribution of the locations of the drones comprises:

10. The method for resource allocation for multi-beam drone cooperative communication according to claim 1, characterized in that said alternating optimization algorithm comprises the following steps:

initializing the number of drones, the user position and the power of each beam;

the association relation between the wave beam and the user set is calculated according to an improved whale optimization algorithm and the network total power consumption of the unmanned aerial vehicle is calculated;

fixing the incidence relation and the position of the unmanned aerial vehicle, optimizing power distribution according to an interior point method, calculating the total power consumption of the network and obtaining a reduction value of the total power consumption of the network;