CN117614520A

CN117614520A - Method for optimizing large-scale MIMO (multiple input multiple output) resources by removing cells based on unmanned aerial vehicle-satellite cooperation

Info

Publication number: CN117614520A
Application number: CN202410091292.8A
Authority: CN
Inventors: 赖海光; 石锋; 孙峰; 张超; 潘奇
Original assignee: Nanjing Kongwei Communication Technology Co ltd
Current assignee: Nanjing Kongwei Communication Technology Co ltd
Priority date: 2024-01-23
Filing date: 2024-01-23
Publication date: 2024-02-27
Anticipated expiration: 2044-01-23
Also published as: CN117614520B

Abstract

The invention discloses a method for optimizing a large-scale MIMO resource for removing cells based on unmanned aerial vehicle-satellite cooperation, which comprises the steps of collecting channel model data from a satellite downlink, preprocessing the channel model data, and constructing a downlink communication transmission model of a large-scale MIMO system for removing cells based on unmanned aerial vehicle-satellite cooperation; aiming at a downlink communication transmission model, constructing a cell-removing large-scale MIMO resource optimization model based on unmanned aerial vehicle-satellite cooperation, and obtaining a non-convex optimization problem; converting the non-convex optimization problem into a preset number of sub-problems by adopting a block coordinate descent method and a continuous convex optimization method in sequence, alternately obtaining the approximate solution of each sub-problem, and obtaining the optimal approximate solution based on the approximate solution of each sub-problem; and calculating and recording performance indexes based on the optimal approximate solution and the downlink communication transmission model. The invention increases the coverage of the user through the optimization method, reduces cross-layer interference and meets the requirement of the rapid increase of communication requirements under special conditions.

Description

Method for optimizing large-scale MIMO (multiple input multiple output) resources by removing cells based on unmanned aerial vehicle-satellite cooperation

Technical Field

The invention belongs to the field of satellite communication, and discloses a method for optimizing a large-scale MIMO resource for removing cells based on unmanned aerial vehicle-satellite cooperation.

Background

Recent studies have shown that de-cellular large-scale multiple-input multiple-output (Cell-Free Massive MIMO) technology has become one of the key research topics for modern networks due to its higher spectral efficiency, higher energy efficiency, and better spatial diversity. However, with the rapid development of wireless communication technology, the demand for supporting mass devices and higher data throughput for the wide area internet of things (IoT) has further increased, and conventional cellular (Massive MIMO) technology has prevented its practical application due to the large number of long cables between each Access Point (AP) and a Central Processing Unit (CPU). Thus, establishing a satellite-assisted air-to-ground network supporting large-scale access is an effective way to solve the current dilemma for a long time, but considering that a satellite network directly provides communication services is a great challenge at present due to limited spectrum resources and communication rates and extremely large delays that are fixed end-to-end.

In the prior art, the unmanned aerial vehicle and the satellite are applied to a cellular large-scale MIMO system, the unmanned aerial vehicle is regarded as a special user, or the unmanned aerial vehicle is assumed to be fixed and has no relation with low propagation, in fact, when the unmanned aerial vehicle is deployed, the height and horizontal coordinates of the unmanned aerial vehicle should be adjusted according to the position of the user, meanwhile, the propagation percentage of the line of sight (LoS) is closely related to the unmanned aerial vehicle, so that the influence of the unmanned aerial vehicle and the satellite is not comprehensively considered in the traditional cellular large-scale communication, so that a blind area exists, and the traditional cellular large-scale communication system needs to use the same time-frequency resource to provide service for a large number of user terminals, so that the shortage and the waste of spectrum resources are caused, and the severity of cross-layer interference and the influence of the cross-layer interference are serious.

Disclosure of Invention

The invention aims to: the method for optimizing the large-scale MIMO resources for removing the cells based on unmanned aerial vehicle-satellite cooperation is provided to solve the problems in the prior art.

The technical scheme is as follows: a method for optimizing a large-scale MIMO resource for removing cells based on unmanned aerial vehicle-satellite cooperation comprises the following steps:

s1, collecting channel model data from a satellite downlink, preprocessing the channel model data, and constructing a downlink communication transmission model of a cellular-removing large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation, wherein the downlink communication transmission model is used for a communication system consisting of a satellite transmitter, a satellite, UAV, CPU, AP and a user;

s2, constructing an objective function and constraint conditions aiming at a downlink communication transmission model, wherein the objective function is maximizing the minimum rate of a user, the constraint conditions are the maximum transmitting power, the maximum cross-layer interference, the maximum communication coverage and the maximum deployment distance of an AP and an UAV, constructing a large-scale MIMO resource optimization model for removing cells based on unmanned aerial vehicle-satellite cooperation, and obtaining a non-convex optimization problem based on the constraint conditions;

s3, sequentially adopting a block coordinate descent method and a continuous convex optimization method, converting the non-convex optimization problem into a preset number of sub-problems, alternately obtaining the approximate solution of each sub-problem, and obtaining the optimal approximate solution based on the approximate solution of each sub-problem;

S4, based on the optimal approximate solution and the downlink communication transmission model, simulating downlink communication of the honeycomb-removed large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation, and calculating and recording performance indexes.

According to one aspect of the present application, the step S1 specifically includes:

s11, receiving channel model data of a satellite downlink, including channel gain, channel phase and channel delay, based on a bidirectional communication link between a satellite transmitter and a satellite;

s12, preprocessing the channel model data, including removing noise, compensating phase offset and correcting channel delay, and improving the accuracy and reliability of the channel model data;

s13, constructing a downlink communication transmission model of the cellular-removing large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation according to the preprocessed channel model data, wherein the downlink communication transmission model comprises a satellite transmitter-satellite-UAV-CPU-AP-user communication transmission model, a CPU-UAV-user communication transmission model, a satellite transmitter-satellite-UAV-user communication transmission model and a satellite transmitter-satellite-user communication transmission model.

According to one aspect of the present application, the step S2 specifically includes:

s21, defining variables and parameters in a cell-removing large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation, wherein the variables comprise a user association matrix, a power allocation matrix and a resource allocation matrix, and the parameters comprise maximum transmitting power, maximum cross-layer interference, maximum communication coverage and maximum deployment distance;

S22, constructing an objective function of the cellular-removing large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation according to variables and parameters and a downlink communication transmission model, maximizing a minimum rate of a user, and defining constraint conditions of the cellular-removing large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation, wherein the constraint conditions comprise maximum transmitting power constraint, maximum cross-layer interference constraint, maximum communication coverage constraint and maximum deployment distance constraint of an AP and an UAV;

s23, constructing a large-scale MIMO resource optimization model for removing the honeycomb based on unmanned aerial vehicle-satellite cooperation aiming at the objective function and the constraint condition, and obtaining the non-convex optimization problem.

According to one aspect of the present application, the step S3 specifically includes:

s31, decomposing a non-convex optimization problem into convex optimization sub-problems related to a user association matrix, a power allocation matrix and a resource allocation matrix by adopting a block coordinate descent method;

s32, linearizing the convex optimization sub-problem by adopting a continuous convex optimization method, converting the convex optimization sub-problem into a convex optimization problem which is easy to solve, solving by adopting an interior point method to obtain at least two approximate solutions of the convex optimization sub-problem, and repeating the process until convergence;

and S33, carrying out weight combination on at least two approximate solutions of the convex optimization sub-problem to obtain an optimal approximate solution of the non-convex optimization problem.

According to one aspect of the present application, the step S4 specifically includes:

s41, determining parameters in a cell-removing large-scale MIMO system of unmanned aerial vehicle-satellite cooperation according to an optimal approximate solution, wherein the parameters comprise a user association matrix, a power distribution matrix, a resource distribution matrix, a channel gain, a channel phase and a channel delay;

s42, inputting the determined parameters into a downlink communication transmission model, and performing simulation calculation on signals in a cellular large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation, wherein the signals comprise satellite transmitter-satellite signals, CPU-AP signals, AP-user signals, CPU-UAV signals and UAV-user signals;

s43, calculating performance indexes including user rate, system throughput, system coverage rate, system interference and system energy consumption according to signals in the cell-removing large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation;

s44, drawing a chart according to the performance index, and analyzing the large-scale MIMO resource optimization method for removing the honeycomb based on unmanned aerial vehicle-satellite cooperation.

According to one aspect of the present application, maximizing the user minimum rate is specifically:

the downlink adopts conjugate beam forming to send signals to the user k, and the mth AP transmission signal is:

x _ik =∑ _k∈Ki sqrt（p _mk ）g _mk d _k

Wherein p is _mk D is the transmission power of the AP to the user downlink _k For signals transmitted to the user, E { |d is satisfied _k ｜ ² ｝=1，g _mk Representing the channel gain between the AP and the user,

the received signals of the user k under the combination of the user schedule are as follows:

y _k =∑ _i ^M∪U ξ _ik g _ik x _ik +ω _k

wherein x is _ik Represents the signal, ω, from the AP _k Is Gaussian white noise meeting normal distribution and uses binary variable xi _k =｛ξ _mk ，ξ _uk Respectively representing the scheduling conditions of users, defining a user scheduling matrix xi epsilon C ^K×（U+1） C is complex set symbol, K is the number of ground users, U is the number of ground unmanned aerial vehicles, M is the number of ground AP points, and binary variable xi is the number of ground AP points if user K is served by the ground AP points _mk The coefficient is 1, otherwise 0,

the user achievable rates for the services provided by the terrestrial APs are:

R _k ^cf =log ₂ （1+ξ _mk ∑ _m=1 ^M （sqrt（p _mk ）g ² _mk ） ² /（∑ _i=1，i≠k ^K ξ _mi ∑ _m=1 ^M （sqrt（p _mi ）g ² _mi ） ² +σ ² ））

wherein sigma ² For the additive white gaussian noise power,

when the unmanned plane u hovers at a fixed position, the user k reachable rate in the communication range is:

R _k ^u =log ₂ （1+ξ _uk ρ _uk h _u，k /（∑ _i=1，i≠k ^K ∑ _u=1 ^U ξ _ui ρ _ui h _u，i +σ ² ））

wherein ρ is _uk Transmitting power for serving unmanned aerial vehicle to user, h _u，k For the height difference between the user and the drone,

the overall reception rate of the user is:

R _k = R _k ^cf + R _k ^u-s

wherein R is _k ^u-s The rate of user k servicing the unmanned satellite.

According to one aspect of the application, the constraint on maximum transmit power is specifically:

the transmission power of the AP and the UAV cannot exceed the maximum transmission power of the AP and the UAV by the resource limitation of the cellular massive MIMO system, and the power distribution parameters of the AP and the UAV need to meet the following constraint:

∑ _k=1 ^K ξ _mk ∑ _m=1 ^M P _mk ＜P _max

∑ _k=1 ^K ξ _uk ρ _uk ＜ρ _max

Wherein P is _max Maximum power supplied to the AP ρ _max Maximum power provided for the drone.

According to one aspect of the application, the problem of non-convex optimization of the honeycomb-removing large-scale MIMO based on unmanned aerial vehicle-satellite cooperation is specifically that:

max _{η，（x，y）} _Шk minR _k =R _k ^cf +R _k ^u-s

C ₁ :∑ _M∪U ξ _mk +ξ _uk ≤1

C ₂ : ξ _mk 、ξ _uk ∈｛0,1｝

C ₃ :‖q _j -q _u ‖≤S _af ,u≠j

s.t. C ₄ ：∑ _k=1 ^K ξ _mk p _mk ＜P _max

C ₅ : ∑ _k=1 ^K ξ _uk ρ _uk ＜ρ _max

C ₆ :∑ _u=1 ^U ∑ _k=1 ^K ξ _uk ρ _sk g _uk ≤ζ _t

wherein ζ _t Interference threshold, q, tolerated for users served by satellite _u Is the horizontal position of the unmanned aerial vehicle, ρ _sk Transmitting power g for serving a satellite to a user _uk Is the channel gain between the drone and the user.

According to one aspect of the present application, step S33 is specifically:

s331, calculating objective function values of each approximate solution according to at least two approximate solutions of the convex optimization sub-problem, and normalizing the objective function values;

s332, assigning corresponding weights to the approximate solutions based on the normalized objective function values;

s333, linearly combining all the approximate solutions according to the weights to obtain the optimal approximate solution of the non-convex optimization problem.

According to one aspect of the application, the method further comprises the step of solving the non-convex optimization problem by adopting a genetic algorithm to obtain an optimal solution, wherein the method specifically comprises the following steps:

s3a, converting parameters of the non-convex optimization problem into binary strings or integer strings, using a random number generator to generate preset initial solutions based on constraint conditions and using the preset initial solutions as parameter values of preset individuals to form an initialization population;

S3b, calculating the fitness of each individual in the initialized population based on the chromosome coding and the objective function to obtain a fitness function;

s3c, sorting the initial solutions based on the fitness function, and selecting a preset initial solution with the top ranking;

s3d, based on constraint conditions, performing crossover and mutation operations on the selected initial solution to generate a new solution;

s3e, rechecking and adjusting the parameter values of the individuals according to the new solution to form a new population, and repeating the steps S3b to S3d until the preset iteration standard is met, so as to obtain the optimal solution.

The beneficial effects are that: the invention provides the method for optimizing the honeycomb-removing large-scale MIMO resource based on the unmanned aerial vehicle-satellite cooperation, comprehensively considers the influence of the unmanned aerial vehicle and the satellite, increases the coverage of users, improves the utilization of system resources, reduces cross-layer interference and meets the requirement of the rapid increase of communication requirements under special conditions.

Drawings

FIG. 1 is a flow chart of the present invention.

Fig. 2 is a flowchart of step S1 of the present invention.

Fig. 3 is a flowchart of step S2 of the present invention.

Fig. 4 is a flowchart of step S3 of the present invention.

Fig. 5 is a flowchart of step S4 of the present invention.

Fig. 6 is a schematic diagram of an aerospace-ground integrated massive MIMO system with massive de-cellular.

Fig. 7 is a flowchart of a resource scheduling algorithm of the unmanned aerial vehicle-satellite assisted space-earth integrated large-scale cell-removing large-scale MIMO system.

Detailed Description

The traditional large-scale communication system for removing the cellular has the problems of resource shortage, cognitive limitation, coverage blind areas and the like, and an Unmanned Aerial Vehicle (UAV) has absolute advantages in terms of rapid deployment, controllable maneuverability, low cost and high probability of air-to-ground link line of sight, and is often used as an air communication auxiliary tool for rapid service recovery after partial or all infrastructure damage caused by natural disasters and communication congestion areas. Therefore, from the standpoint of network comprehensive performance and cost, the auxiliary cellular-removal large-scale MIMO system consisting of the unmanned aerial vehicle and the satellite has a great application prospect. The network can effectively improve challenges brought by a large number of long cables in a cellular large-scale MIMO system, and can also improve the cognitive limitation of the system, so that no blind spot coverage of users is realized. The invention provides a method for optimizing a large-scale MIMO resource for removing cells based on unmanned aerial vehicle-satellite cooperation, which is shown in figure 1 specifically and comprises the following steps:

s1, collecting channel model data from a satellite downlink, preprocessing the channel model data, and constructing a downlink communication transmission model of a cellular large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation, wherein the downlink communication transmission model comprises a satellite transmitter-satellite-UAV-CPU-AP-user communication transmission model, a CPU-UAV-user communication transmission model, a satellite transmitter-satellite-UAV-user communication transmission model and a satellite transmitter-satellite-user communication transmission model;

The step S1, as shown in fig. 2, includes:

s13, constructing a downlink communication transmission model of the honeycomb-removing large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation according to the preprocessed channel model data.

In a further embodiment, an assisted system consisting of a plurality of single antenna Unmanned Aerial Vehicles (UAVs) and a satellite serves a square area of ground wireless communication in an unmanned aerial vehicle-satellite assisted space-ground integrated massive-scale-de-cellular MIMO system. The ground is provided with a plurality of Access Points (APs) which are connected to a Central Processing Unit (CPU) and cooperate to provide services for users through ideal backhaul links, and the unmanned aerial vehicle accesses the CPU network in a wireless mode and shares system information. Because the spectrum resources are limited, in a single time slot, the satellite and any UAV share the spectrum resources, and different UAVs operate on different frequency bands, so that cell interference cannot exist among different UAVs, but the interference can be generated on satellite users k using the same spectrum resources, meanwhile, most users in a scene are pedestrians, the situation can be regarded as stationary, and data can be transmitted in real time between a satellite transmitter connected with the satellite and a CPU and connected with the users through an AP access point. In special cases, no one has access to the CPU network to serve the range that the AP access point cannot serve. For the modeling of the unique channel environment of satellite communication, a communication transmission system of a MIMO system of a satellite transmitter-satellite-UAV-CPU-AP-user and a communication transmission system of the CPU-UAV-user are established, and even the communication transmission system of the satellite transmitter-satellite-UAV-user and the satellite transmitter-satellite-user is realized.

The large-scale MIMO communication system for removing the honeycomb based on unmanned aerial vehicle-satellite cooperation aims at the requirement of space-ground integration in the 6G development, and solves the problems of large-scale fading, small-scale fading and the like caused by pitch angle in places where a plurality of AP access points cannot cover or by applying UAV communication.

Further, step S11 specifically includes:

s11a, receiving channel model data of a satellite downlink based on a two-way communication link between a satellite transmitter and a satellite;

s11b, periodically collecting position, speed and acceleration information of the satellite and the unmanned aerial vehicle by a CPU, and predicting and tracking the motion states of the satellite and the unmanned aerial vehicle by a Kalman filtering method according to the collected information;

s11c, according to the motion states of the satellite and the unmanned aerial vehicle, predicting and tracking time variability and non-stationarity of the channel model data by adopting a Doppler frequency shift method, and dynamically updating and correcting the channel model data;

and S11d, carrying out channel estimation by adopting a least square method according to the updated and corrected channel model data to obtain channel gain, channel phase and channel delay.

The uncertainty and the variability of channel model data caused by high-speed movement of the satellite and the unmanned aerial vehicle and complex environment can be effectively processed by predicting and tracking the movement states of the satellite and the unmanned aerial vehicle, so that the channel tracking effect and the communication stability of the large-scale MIMO system for going to the honeycomb based on the cooperation of the unmanned aerial vehicle and the satellite are improved; by predicting and tracking the time variability and the non-stationarity of the channel model data and dynamically updating and correcting the channel model data, the frequency offset and the time delay variation of the channel model data caused by the relative motion and the signal propagation characteristics of the satellite and the unmanned aerial vehicle can be effectively processed, so that the channel adaptability and the communication efficiency of the large-scale MIMO system for removing the honeycomb based on the unmanned aerial vehicle-satellite cooperation are improved.

S2, constructing a honeycomb-removing large-scale MIMO resource optimization model based on unmanned aerial vehicle-satellite cooperation aiming at a downlink communication transmission model, wherein the objective function is the minimum rate of a maximized user, and constraint conditions are the maximum transmitting power, the maximum cross-layer interference, the maximum communication coverage and the maximum deployment distance of an AP and a UAV, so that a non-convex optimization problem is obtained;

the step S2, as shown in fig. 3, includes:

s22, constructing an objective function of the cellular-removing large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation according to the variables and parameters and a downlink communication transmission model, maximizing a minimum rate of a user, and defining constraint conditions of the cellular-removing large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation, wherein the constraint conditions comprise maximum transmitting power constraint, maximum cross-layer interference constraint, maximum communication coverage constraint and maximum deployment distance constraint of an AP and an UAV;

The method for optimizing the large-scale MIMO resources by removing the cells based on unmanned aerial vehicle-satellite cooperation is provided, and an optimization model is built. The method aims at maximizing the minimum rate of the user and meets the constraints of maximum transmitting power, maximum cross-layer interference, maximum communication coverage and maximum deployment distance.

The system maximizes the minimum data transmission rate among all users by optimizing resource allocation and parameter configuration in order to ensure communication quality at the user side, so that each user can obtain higher communication rate in the whole system, and user experience and satisfaction are improved. By adjusting the policies of communication access point layout, power allocation, and resource allocation, the system can optimize channel quality and signal transmission efficiency, thereby maximizing the minimum rate of users.

Further, the maximizing the user minimum rate is specifically:

since the channel conditions are known, the downlink transmits a signal to the user k by using conjugate beam forming, and the mth AP transmits the signal:

x _ik =∑ _k∈Ki （sqrt p _mk ）g _mk d _k

Wherein p is _mk D is the transmission power of the AP to the user downlink _k For signals transmitted to the user, E { |d is satisfied _k ｜ ² ｝=1，g _mk Representing the channel gain between the AP and the user, i is a variable,

y _k =∑ _i ^M∪U ξ _ik g _ik x _ik +ω _k

the user achievable rates for the services provided by the terrestrial APs are:

R _k ^cf =log ₂ （1+ξ _mk ∑ _m=1 ^M （（sqrt p _mk ）g ² _mk ） ² /（∑ _i=1，i≠k ^K ξ _mi ∑ _m=1 ^M （（sqrt p _mi ）g ² _mi ） ² +σ ² ））

wherein sigma ² For the additive white gaussian noise power,

because the over-the-air AP only provides business services for users in the communication coverage area, and does not interfere with other user signals, when the unmanned aerial vehicle hovers at a fixed position, the user reachable rate in the communication coverage area is as follows:

the overall reception rate of the user is:

R _k = R _k ^cf + R _k ^u-s

wherein R is _k ^u-s The rate of user k servicing the unmanned satellite.

The maximum transmit power constraint means that by reasonably adjusting and limiting the power of the communication device, the system can achieve optimal power utilization efficiency and effectively control the energy consumption of the device, and meanwhile, the constraint also helps to avoid excessive interference and spectrum waste, and improves the overall communication performance and reliability. In the resource allocation and optimization process, the system considers the maximum transmitting power constraint, and dynamically adjusts and manages the power output of the communication equipment according to the specific situation so as to meet the system requirement and the user requirement.

Further, the constraint condition of using the maximum transmitting power is specifically:

∑ _k=1 ^K ξ _mk ∑ _m=1 ^M P _mk ＜P _max

∑ _k=1 ^K ξ _uk ρ _uk ＜ρ _max

The maximum cross-layer interference constraint describes the requirements of the system to control the interference level between different communication layers during resource allocation and communication parameter configuration. By ensuring mutual cooperation between the communication nodes, the influence of interference on the communication quality is reduced. Cross-layer interference refers to the phenomenon of mutual interference between different communication layers, including a physical layer, a data link layer, a network layer, etc., which may cause degradation of signal quality, increase of bit error rate, and slow down of data transmission rate. To avoid or reduce cross-layer interference, the system may take a number of strategies including efficient spectrum allocation, power control, scheduling algorithms, interference suppression techniques, and the like. Through reasonable resource management and optimization algorithm, the system can dynamically adjust parameters of different communication layers to minimize cross-layer interference and improve overall communication performance. This helps to provide a stable, high quality communication service while maximizing user experience and satisfaction.

The maximum communication coverage constraint is such that the system ensures that users within the transmission range of the communication device are able to receive a stable and high quality signal. The communication coverage of a communication device depends on a number of factors including the transmit power of the device, the antenna gain, the transmission frequency, and the environmental conditions, among others. The system will optimize and manage these factors to expand the coverage of the communication device, maintaining the stability and reliability of signal transmission. In the resource allocation and layout process, the system considers the deployment position, antenna direction, antenna height and other factors of the equipment so as to extend the communication coverage to the greatest extent. Meanwhile, the system may also adopt methods such as signal enhancement technology, reasonable signal transmission protocol and scheduling algorithm, etc. to improve the communication quality and use experience of users in the coverage area.

The system is to control the maximum deployment distance between the communication devices to ensure stable and reliable communication connection between the devices. The maximum deployment distance of a communication device depends on a number of factors including transmission technology, transmission frequency, signal propagation characteristics, and device deployment environment. The system will optimize and manage these factors to control the deployment distance between devices and maintain the stability and reliability of the communication connection. In the resource allocation and deployment process, the system can consider the deployment position, antenna direction, antenna height and other factors of the communication equipment so as to control the maximum deployment distance. Meanwhile, the system may also adopt methods such as signal enhancement technology, reasonable signal transmission protocol and scheduling algorithm, etc. so as to improve the communication quality and use experience between the devices.

The constraint conditions are maximum transmitting power constraint, maximum cross-layer interference constraint, maximum communication coverage constraint and maximum deployment distance constraint respectively ensure energy consumption and electromagnetic wave radiation control, ensure mutual cooperation among communication nodes, ensure that users in the communication coverage area can receive stable and high-quality signals, and ensure signal transmission efficiency and communication quality among devices.

Further, for the unmanned aerial vehicle satellite auxiliary system designed for the traditional honeycomb-removing large-scale MIMO system, the aim is to optimize the user scheduling, the UAV track and the transmitting power in a combined way, maximize the minimum average rate of the user, expand the system coverage area and obtain the following optimization problems:

max _{η，（x，y）} _Шk minR _k =R _k ^cf +R _k ^u-s

C ₁ :∑ _M∪U ξ _mk +ξ _uk ≤1

C ₂ : ξ _mk 、ξ _uk ∈｛0,1｝

C ₃ :‖q _j -q _u ‖≤S _af ,u≠j

s.t. C ₄ ：∑ _k=1 ^K ξ _mk p _mk ＜P _max

C ₅ : ∑ _k=1 ^K ξ _uk ρ _uk ＜ρ _max

C ₆ :∑ _u=1 ^U ∑ _k=1 ^K ξ _uk ρ _sk g _uk ≤ζ _t

wherein R is _k ^cf For the rate of user k served by the AP, R _k ^u-s Representing the rate, ζ, of user k served by the unmanned aerial vehicle satellite _t Interference threshold, q, tolerated for users served by satellite _u Is the horizontal position of the unmanned aerial vehicle, ρ _sk Transmitting power g for serving a satellite to a user _uk For the channel gain between the drone and the user, the overall reception rate R of the user _k User reachable rate R that can be broken down to be served by a terrestrial AP _k ^cf Sum R of user reachable rates in communication range with unmanned aerial vehicle _k ^u-s J is a variable.

the step S3, as shown in fig. 4, includes:

s32, linearizing the convex optimization sub-problem by adopting a continuous convex optimization method, converting the convex optimization sub-problem into a convex optimization problem which is easier to solve, solving by adopting an interior point method to obtain at least two approximate solutions of the convex optimization sub-problem, and repeating the process until convergence;

And introducing auxiliary variables into the complex optimization objective function, and decomposing the original problem into a plurality of sub-problems which can be optimized alternately by utilizing block coordinate descent. Aiming at the optimization problem, adopting a continuous convex optimization (Successive Convex Approximation, SCA) technology, converting the sub-problem into a convex optimization problem and sequentially solving the convex optimization problem to finally obtain the approximate optimal solution of the original problem.

In a further embodiment, to solve constraint C ₁ Is approximated by a continuous convex optimization method _k Specifically, in each iteration, R' _k Approximating ζ at a given local point by a more tractable function _r （ξ _mk ^r ，ξ _uk ^r ) Defined as the result of the user association in the r-th iteration. The global upper bound is extended by any concave function at any point by its first order taylor, for a given local point ζ _r There is a convex upper bound.

Further, step S33 specifically includes:

Therefore, the solving precision and stability of the optimization problem can be effectively improved, and the resource allocation and management effects and the communication quality of the large-scale MIMO system for removing the honeycomb based on unmanned aerial vehicle-satellite cooperation are improved.

The step S4, as shown in fig. 5, includes:

s42, inputting the determined parameters into a downlink communication transmission model, and performing simulation calculation on signals in a cellular-removing large-scale MIMO system based on unmanned aerial vehicle-satellite cooperation, wherein the signals comprise satellite transmitter-satellite signals, satellite-CPU signals, CPU-AP signals, AP-user signals, CPU-UAV signals and UAV-user signals;

In a further embodiment, according to the known deployment defects of the users and the ground access points, the coverage constraint, the maximum power constraint and the cross-layer interference constraint of each communication access point are considered to maximize the minimum rate of the users, and a hybrid resource allocation model combining user association, power allocation and unmanned plane placement is established. Based on a block coordinate descent method and a continuous convex optimization method, the original non-convex optimization problem is converted into 3 sub-problems, sub-problem approximate solutions are obtained alternately, and finally the optimal approximate solution of the original problem is obtained. Simulation results show that the proposed method can remarkably improve the communication coverage of the system, reasonably arrange the resource placement of the system and improve the average throughput of users, and specifically comprises the following steps:

Step one, collecting a channel model from a downlink of a geosynchronous orbit satellite, and performing preliminary estimation on each large-scale fading of the channel gain. And establishing a downlink communication transmission system of the unmanned aerial vehicle-satellite assisted cellular-removal large-scale MIMO system.

And secondly, considering coverage constraint, maximum power constraint and cross-layer interference constraint of each communication access point to maximize the minimum rate of users, and establishing a hybrid resource allocation model of joint user association, power allocation and geosynchronous satellite.

And thirdly, establishing an optimization problem by taking the minimum rate of the maximized user as a target and meeting the maximum transmitting power constraint, the maximum cross-layer interference constraint, the maximum communication coverage constraint and the maximum deployment distance constraint.

And step four, converting the original non-convex optimization problem into a plurality of sub-problems based on a block coordinate descent method and a continuous convex optimization method, alternately obtaining sub-problem approximate solutions, and finally obtaining the optimal approximate solution of the original problem.

And fifthly, comparing the gap between the proposed method and the traditional honeycomb-removed large-scale MIMO system in terms of data such as performance coverage rate, average user throughput and the like according to simulation model results, and finally obtaining simulation results.

Furthermore, a method based on cryptography or blockchain is adopted to carry out security and privacy protection on the large-scale MIMO system for removing the honeycomb of unmanned aerial vehicle-satellite cooperation.

Further, a deep learning technology is used for constructing a downlink communication transmission model of the cellular-removing large-scale MIMO system of the unmanned plane-satellite cooperation. The deep learning technology is an artificial intelligence technology based on a multi-layer neural network, can learn complex characteristics and rules from a large amount of data, and improves the expression capacity and generalization capacity of a model. A convolutional neural network or a cyclic neural network is used for establishing a channel model, a deep reinforcement learning or deep evolution algorithm is used for optimizing resource allocation and scheduling strategies, and a deep generation model or a deep self-adaptive model is used for adapting to dynamic communication environment and user requirements. The deep learning technology is used for improving the performance and efficiency of a downlink communication transmission model of the cellular-removing large-scale MIMO system of the unmanned aerial vehicle-satellite cooperation, reducing channel estimation errors and feedback delay, improving the system capacity and robustness, and enhancing the intellectualization and autonomy of the system.

In a further embodiment, as shown in fig. 6, a schematic diagram of an air-to-ground integrated massive de-cellular massive MIMO system is provided, a communication transmission system of the MIMO system by the satellite transmitter-satellite-UVA-CPU-AP-user and a communication transmission system of the CPU-UAV-user are established for unique channel environment modeling of satellite communication, and even the satellite transmitter-satellite-UAV-user and the satellite transmitter-satellite-user are implemented.

In a further embodiment of the present application,

constructing a channel model of unmanned aerial vehicle satellite cooperation of unmanned aerial vehicle-satellite cooperation, taking the motion characteristics of the unmanned aerial vehicle into consideration, adopting a method based on a geometric random process, constructing a space channel model of unmanned aerial vehicle satellite cooperation, and describing the channel state transition probability and the channel gain change rule between the unmanned aerial vehicle and the satellite;

the orbit characteristics of satellites are considered, a space channel model of unmanned aerial vehicle satellite cooperation is established by adopting an orbit dynamics-based method, and channel state transition probability and channel gain change rules between the satellites and a ground base station are described.

And (3) establishing a space channel model of unmanned aerial vehicle satellite cooperation by adopting an angle distribution-based method in consideration of the relative position relationship between the unmanned aerial vehicle and the satellite, and describing the channel direction angle and the channel correlation between the unmanned aerial vehicle and the satellite.

Constructing a frequency domain channel model of unmanned aerial vehicle satellite cooperation:

analyzing the channel frequency selectivity of unmanned aerial vehicle satellite cooperation, establishing a frequency domain channel model of unmanned aerial vehicle satellite cooperation by adopting a method based on multipath fading, and describing the channel frequency response and the channel bandwidth between the unmanned aerial vehicle and the satellite;

analyzing the Doppler effect of the channel coordinated with the unmanned aerial vehicle satellite, establishing a frequency domain channel model coordinated with the unmanned aerial vehicle satellite by adopting a Doppler frequency shift-based method, and describing the Doppler frequency shift and the coherence time of the channel between the unmanned aerial vehicle and the satellite;

Analyzing the time variability of the channel cooperated with the unmanned aerial vehicle satellite, establishing a frequency domain channel model cooperated with the unmanned aerial vehicle satellite by adopting a Kalman filtering-based method, and describing the time-varying characteristics of the channel between the unmanned aerial vehicle and the satellite and a channel prediction method.

In another embodiment of the present application, the process of the network topology model of unmanned aerial vehicle satellite collaboration is further:

the network hierarchical structure of unmanned aerial vehicle satellite cooperation is analyzed and considered, a network architecture model of unmanned aerial vehicle satellite cooperation is established by adopting a method based on a software defined network, and separation and cooperation of a network control layer and a network data layer between the unmanned aerial vehicle and the satellite are described.

Analyzing and considering network function division of unmanned aerial vehicle satellite cooperation, establishing a network architecture model of unmanned aerial vehicle satellite cooperation by adopting a network function virtualization-based method, and describing virtualization and migration of network functions between an unmanned aerial vehicle and satellites.

Analyzing a network access mode of unmanned aerial vehicle satellite cooperation, establishing a network architecture model of unmanned aerial vehicle satellite cooperation by adopting a non-orthogonal multiple access-based method, and describing a network access mode and a resource allocation strategy between the unmanned aerial vehicle and a satellite.

Analyzing network energy consumption optimization of unmanned aerial vehicle satellite cooperation, and establishing a network optimization model of unmanned aerial vehicle satellite cooperation by adopting a game theory-based method to describe network energy consumption optimization problems and Nash equilibrium solutions between the unmanned aerial vehicle and satellites.

And analyzing network capacity optimization of unmanned aerial vehicle satellite cooperation, and establishing a network optimization model of unmanned aerial vehicle satellite cooperation by adopting a graph theory-based method to describe network capacity optimization problems and maximum flow minimum theorem between the unmanned aerial vehicle and the satellite.

Analyzing network delay optimization of unmanned aerial vehicle satellite cooperation, establishing a network optimization model of unmanned aerial vehicle satellite cooperation by adopting a queuing theory-based method, and describing network delay optimization problems and steady state probability distribution between the unmanned aerial vehicle and the satellite.

In another embodiment of the present application, the unmanned aerial vehicle satellite cooperative de-cellular massive MIMO system performance analysis includes:

and analyzing the signal-to-noise ratio of the unmanned aerial vehicle satellite-cooperative cellular massive MIMO system by adopting a method based on a random matrix theory, and deducing an asymptotic expression of the signal-to-noise ratio of the unmanned aerial vehicle satellite-cooperative cellular massive MIMO system.

And analyzing the signal-to-noise ratio of the unmanned aerial vehicle satellite-cooperative cellular massive MIMO system by adopting a method based on a random geometric model, and deducing an accurate expression of the signal-to-noise ratio of the unmanned aerial vehicle satellite-cooperative cellular massive MIMO system.

And analyzing the signal-to-noise ratio of the unmanned aerial vehicle satellite-cooperative cellular massive MIMO system by adopting a non-ideal hardware model-based method, and deducing an approximate expression of the signal-to-noise ratio of the unmanned aerial vehicle satellite-cooperative cellular massive MIMO system.

In another embodiment of the present application, the error rate analysis of the cell-free massive MIMO system with unmanned aerial vehicle satellite cooperation includes:

and analyzing the error rate of the cell-removing large-scale MIMO system of the unmanned aerial vehicle satellite cooperation by adopting a method based on a random matrix theory, and deducing an asymptotic expression of the error rate of the cell-removing large-scale MIMO system of the unmanned aerial vehicle satellite cooperation.

And analyzing the error rate of the unmanned aerial vehicle satellite-cooperated cell-removing large-scale MIMO system by adopting a method based on a random geometric model, and deducing an accurate expression of the error rate of the unmanned aerial vehicle satellite-cooperated cell-removing large-scale MIMO system.

And analyzing the error rate of the unmanned aerial vehicle satellite-cooperated cell-removing large-scale MIMO system by adopting a method based on a non-ideal hardware model, and deducing an approximate expression of the error rate of the unmanned aerial vehicle satellite-cooperated cell-removing large-scale MIMO system.

In another embodiment of the present application, a cell-free massive MIMO system capacity analysis for unmanned aerial vehicle satellite cooperation includes:

and analyzing the capacity of the large-scale MIMO system for removing the honeycomb of the unmanned aerial vehicle satellite cooperation by adopting a method based on a random matrix theory, and deducing an asymptotic expression of the capacity of the large-scale MIMO system for removing the honeycomb of the unmanned aerial vehicle satellite cooperation.

And analyzing the capacity of the large-scale MIMO system for removing the honeycomb of the unmanned aerial vehicle satellite cooperation by adopting a method based on a random geometric model, and deducing an accurate expression of the capacity of the large-scale MIMO system for removing the honeycomb of the unmanned aerial vehicle satellite cooperation.

And analyzing the capacity of the unmanned aerial vehicle satellite cooperative cellular massive MIMO system by adopting a non-ideal hardware model-based method, and deducing an approximate expression of the capacity of the unmanned aerial vehicle satellite cooperative cellular massive MIMO system.

In another embodiment of the present application, an unmanned aerial vehicle satellite-cooperative de-cellular massive MIMO resource optimization algorithm,

defining variables and parameters in a cellular-removing large-scale MIMO system of unmanned aerial vehicle satellite cooperation, wherein the variables comprise a user association matrix, a power distribution matrix and a resource distribution matrix, and the parameters comprise maximum transmitting power, maximum cross-layer interference, maximum communication coverage and maximum deployment distance;

according to the variables and parameters and performance indexes of the unmanned aerial vehicle satellite cooperative cellular-removal large-scale MIMO system, an objective function of the unmanned aerial vehicle satellite cooperative cellular-removal large-scale MIMO system is constructed, the total capacity of the system is maximized, constraint conditions of the unmanned aerial vehicle satellite cooperative cellular-removal large-scale MIMO system are defined, and the constraint conditions comprise maximum transmitting power constraint, maximum cross-layer interference constraint, maximum communication coverage constraint and maximum deployment distance constraint of the AP and the UAV;

Aiming at objective functions and constraint conditions, the problem of optimization of the non-convex optimization is obtained by constructing the problem of optimization of the non-cellular large-scale MIMO resource of unmanned aerial vehicle satellite cooperation.

Honeycomb-removing large-scale MIMO resource optimization algorithm for constructing unmanned aerial vehicle satellite cooperation

Decomposing the non-convex optimization problem into convex optimization sub-problems related to a user association matrix, a power allocation matrix and a resource allocation matrix by adopting a method based on an alternate direction multiplication sub-method; solving the convex optimization sub-problem by adopting a gradient descent method to obtain a local optimal solution of the convex optimization sub-problem, and repeating the process until convergence; and merging the local optimal solutions of the convex optimization sub-problems to obtain a global optimal solution of the non-convex optimization problem.

In another embodiment of the present application, the building of the simulation platform of the cellular-removing massive MIMO system for satellite cooperation of the unmanned aerial vehicle further includes:

and (3) building a physical layer simulation platform of the cellular-removing large-scale MIMO system for unmanned aerial vehicle satellite cooperation by adopting a method based on software defined radio, and realizing software implementation and hardware mapping of a channel model, a network architecture model and a resource optimization model for unmanned aerial vehicle satellite cooperation. A network simulation-based method is adopted to build a network layer simulation platform of the cellular-removing large-scale MIMO system for unmanned aerial vehicle satellite cooperation, and software implementation and data analysis of a network topology model, a network optimization model and a network performance evaluation model for unmanned aerial vehicle satellite cooperation are realized. And an application layer simulation platform of the honeycomb-removed large-scale MIMO system for unmanned aerial vehicle satellite cooperation is built by adopting a hybrid simulation-based method, so that software implementation and user experience test of an application scene model, an application demand model and an application effect evaluation model for unmanned aerial vehicle satellite cooperation are realized.

In another embodiment of the present application, the simulation parameter setting of the cellular-removing massive MIMO system for unmanned aerial vehicle satellite cooperation includes:

according to performance indexes of the large-scale MIMO system for removing the honeycomb of the unmanned aerial vehicle satellite collaboration, simulation parameters of the large-scale MIMO system for removing the honeycomb of the unmanned aerial vehicle satellite collaboration are set, wherein the simulation parameters comprise the number, position, speed, height, service area and the like of unmanned aerial vehicles, the number, orbit, height, coverage area and the like of satellites, the number, position, power, antenna number and the like of APs, the number, position, speed and the like of users.

According to a resource optimization algorithm of the unmanned aerial vehicle satellite cooperative honeycomb removal large-scale MIMO system, setting optimization parameters of the unmanned aerial vehicle satellite cooperative honeycomb removal large-scale MIMO system, wherein the optimization parameters comprise weight of an objective function, a threshold value of a constraint condition, iteration times of the optimization algorithm, convergence conditions and the like.

According to the application scene of the unmanned aerial vehicle satellite cooperative cellular massive MIMO system, setting the application parameters of the unmanned aerial vehicle satellite cooperative cellular massive MIMO system, including application types, application data amounts, application duration, application quality and the like.

In another embodiment of the present application, the simulation result analysis of the cellular-removing massive MIMO system for unmanned aerial vehicle satellite cooperation includes:

And (3) operating a physical layer simulation platform of the cellular-removing large-scale MIMO system with unmanned aerial vehicle satellite cooperation, and recording physical layer simulation results of the cellular-removing large-scale MIMO system with unmanned aerial vehicle satellite cooperation, wherein the physical layer simulation results comprise signal-to-noise ratio, bit error rate, capacity and the like.

And (3) operating a network layer simulation platform of the cellular-removing large-scale MIMO system with unmanned aerial vehicle satellite cooperation, and recording network layer simulation results of the cellular-removing large-scale MIMO system with unmanned aerial vehicle satellite cooperation, wherein the network layer simulation results comprise network energy consumption, network capacity, network delay and the like. And running an application layer simulation platform of the cellular-removing large-scale MIMO system of the unmanned aerial vehicle satellite collaboration, and recording an application layer simulation result of the cellular-removing large-scale MIMO system of the unmanned aerial vehicle satellite collaboration, wherein the application layer simulation result comprises application throughput, application success rate, application satisfaction and the like.

In another embodiment of the present application, the building of the unmanned aerial vehicle satellite-collaborative cellular-removal massive MIMO system experiment platform includes:

the method is characterized in that a physical layer experimental platform of a cellular-removing large-scale MIMO system with unmanned aerial vehicle satellite cooperation is built by adopting a method based on software defined radio, and hardware realization and software control of a channel model, a network architecture model and a resource optimization model with unmanned aerial vehicle satellite cooperation are realized by utilizing real equipment of unmanned aerial vehicles, satellites, APs and users.

Step S412: a network test-based method is adopted to build a network layer experimental platform of the cellular-removing large-scale MIMO system of unmanned aerial vehicle satellite cooperation, and a network test instrument is utilized to realize hardware realization and data acquisition of a network topology model, a network optimization model and a network performance evaluation model of unmanned aerial vehicle satellite cooperation. An application layer experimental platform of the cellular-removing large-scale MIMO system for unmanned aerial vehicle satellite collaboration is built by adopting a method based on user experience, and hardware implementation and user feedback of an application scene model, an application demand model and an application effect evaluation model for unmanned aerial vehicle satellite collaboration are realized by utilizing the real demands of users. According to performance indexes of the large-scale MIMO system for removing the honeycomb of the unmanned aerial vehicle satellite collaboration, experimental parameters of the large-scale MIMO system for removing the honeycomb of the unmanned aerial vehicle satellite collaboration are set, wherein the experimental parameters comprise the number, position, speed, height, service area and the like of unmanned aerial vehicles, the number, orbit, height, coverage area and the like of satellites, the number, position, power, antenna number and the like of APs, the number, position, speed and the like of users. According to a resource optimization algorithm of the unmanned aerial vehicle satellite cooperative honeycomb removal large-scale MIMO system, setting optimization parameters of the unmanned aerial vehicle satellite cooperative honeycomb removal large-scale MIMO system, wherein the optimization parameters comprise weight of an objective function, a threshold value of a constraint condition, iteration times of the optimization algorithm, convergence conditions and the like. According to the application scene of the unmanned aerial vehicle satellite cooperative cellular massive MIMO system, setting the application parameters of the unmanned aerial vehicle satellite cooperative cellular massive MIMO system, including application types, application data amounts, application duration, application quality and the like.

In another embodiment of the application, a physical layer experiment platform of the large-scale MIMO system for removing the honeycomb of the unmanned aerial vehicle satellite cooperation is operated, and physical layer experiment results of the large-scale MIMO system for removing the honeycomb of the unmanned aerial vehicle satellite cooperation are recorded, wherein the physical layer experiment results comprise signal-to-noise ratio, bit error rate, capacity and the like. And (3) operating a network layer experimental platform of the cellular-removing large-scale MIMO system of the unmanned aerial vehicle satellite collaboration, and recording network layer experimental results of the cellular-removing large-scale MIMO system of the unmanned aerial vehicle satellite collaboration, wherein the network layer experimental results comprise network energy consumption, network capacity, network delay and the like. And (3) operating an application layer experimental platform of the cellular-removing large-scale MIMO system of the unmanned aerial vehicle satellite collaboration, and recording an application layer experimental result of the cellular-removing large-scale MIMO system of the unmanned aerial vehicle satellite collaboration, wherein the application layer experimental result comprises application throughput, application success rate, application satisfaction and the like.

Further, a specific algorithm flow of the unmanned aerial vehicle-satellite assisted space-earth integrated large-scale cell removal large-scale MIMO system resource scheduling algorithm is shown in fig. 7, and specifically comprises the following steps:

step a: initializing a binary selection variable ζ of a user ⁰ Horizontal position q of unmanned aerial vehicle ⁰ Maximum power P provided by AP ⁰ Iteration meterA counter r=0, error ψ;

step b: for a given { ζ } ^r ，q ^r ，P ^r Solution due to C ₁ Non-convex optimization problem caused by non-convex constraint, C ₁ R in (a) _k The problem of poor control variable relative to the association factor can be written as two concave functions, and the current optimal solution xi is obtained ^r+1 。

R _k ^cf =log ₂ （∑ _i=1 ^K ξ _mi ∑ _m=1 ^M （sqrt（p _mi ）g ² _mi ） ² +σ ² ）-R＇ _k

Wherein,

R＇ _k = log ₂ （∑ _i=1，i≠k ^K ξ _mi ∑ _m=1 ^M （sqrt（p _mi ）g ² _mi ） ² +σ ² ）

step c: for a given { ζ } ^r+1 ，q ^r ，P ^r Solution to obtain problem optimal solution q ^r+1 Regarding II q _u-s -q _k ‖ ² Is convex, the convex function is the lower bound of the first-order Taylor expansion of the convex function at any point, and the unmanned aerial vehicle deployment result q is given at any r time ^r ，R~ _k，u The lower bound function of (2) is:

wherein A is _K，u ^r 、B _k，u ^r Respectively a first-order coefficient and a constant,

A _K，u ^r =b _k，u （‖q _u -q _k ‖ ² -‖q _u ^r -q _k ‖ ² ）（1/In2）/（H ² +‖q _u ^r -q _k ‖ ² ）（b _k，u + H ² +‖q _u ^r -q _k ‖ ² ）

B _k，u ^r = log ₂ （1+ b _k，u /（H ² +‖q _u ^r -q _k ‖ ² ））

step d: for a given { ζ } ^r+1 ，q ^r+1 ，P ^r Solution to obtain problem optimal solution P ^r+1 At a given iteration result q ^r Lower bound function R _k，u ^lb Is used for solving the problems of the prior art,

max _λ，q λ

C ₁ : R _k ^cf + R _k ^u-s ≥λ，Шk

s.t. C ₂ ：‖q _j -q _u ‖≥S _af ，Шu、j∈U，u≠j

the approximate unmanned aerial vehicle deployment optimization problem is as follows:

max _λq ^r _，q λ _q ^r

C ₁ : R _k ^cf + R _u ^lb ≥λ _q ^r

s.t. C ₂ ：-‖q _j ^r -q _u ^r ‖+2（q _j ^r -q _u ^r ） ^T （q _j -q _u ）≥S _af ，P _m

Шu、j∈U，u≠j

wherein,

λ（ξ，q，P）=minR _k =R _k ^cf +R _k ^u-s

this is a convex optimization problem that can be solved efficiently by a convex optimizer;

step e: the number of iterations r=r+1 is updated, bounded by the iteration accuracy reaching a threshold, i.e. until |λ ^r+1 -λ ^r And c, skipping the step b, wherein |is larger than or equal to psi.

In a further embodiment, in a downlink communication link consisting of an AP, a UAV, a satellite, a user, the information interaction in the link is performed according to the following steps:

firstly, deploying transmission units such as an AP, a UAV, a satellite and the like in a system, and constructing a downlink communication link, wherein the AP, the UAV and the satellite serve users together;

Elements in the system model include APs, UAVs, satellites, and users, where the APs, UAVs, satellites collectively serve users, and where their energy and spectral resources are limited. The set of users is defined as k= {1,2, …, K }, the set of APs is defined as m= {1,2, …, M }, the set of UAVs is defined as u= {1,2, …, U }, the number of satellites is only 1 and the corresponding set is defined as s= {1}.

Establishing a communication model considering multidimensional interference, wherein the model comprises an achievable rate when an AP (access point) serves a user, an achievable rate when a UAV (unmanned aerial vehicle) serves the user, an achievable rate when a satellite serves the user and interference generated in three downlink transmission processes;

the user can only associate with one of the AP, the UAV and the satellite to receive the information transmitted downwards, thus a group of binary variables are introduced to describe the association situation of the user, and the association situation of the kth user is defined as xi _k （ξ _mk 、ξ _uk 、ξ _sk ) Respectively comprises access conditions of AP, UAV and satellite, wherein ζ _mk 、ξ _uk 、ξ _sk Only one of the three can have a value of 1, representing downlink data transmission in the mode, and the other two are 0, representing data transmission not in the mode, so that xi is respectively carried out according to the association condition of the kth user, the AP, the unmanned aerial vehicle and the satellite _mk ∈｛0，1｝、ξ _uk ∈｛0，1｝、ξ _sk E {0,1}, and finally the correlation matrix facing all users is defined as xi E C ^{K×（M+U+1）} Wherein C is a complex set symbol.

If user k is associated with ground APm, binary variable ζ _mk The coefficient is 1, otherwise 0. Specifically, the user's reachable rate is defined by R _km ^c The specific expression is as follows:

R _km ^c =ξ _mk z _mk log ₂ （1+（p _m g _mk /（∑ _i=1，i≠m ^M p _i g _ik +∑ _j=1 ^U p ^～ _j g ^～ _jk +p ^、 _s g ^、 _sk +σ ² ）））

wherein z is _mk Representing downstream bandwidth resources, p, when user k associates with an mth AP access point _m Is the transmission power of APm g _mk Is the channel gain between APm and user k. In the denominator part, the first term is the out-of-cluster interference from other APs, where p _i And g _ik Respectively representing the transmission power of APi and the channel gain between the APi and user k; the second term is the out-of-cluster interference from all drones, where p ^～ _j And g ^～ _jk Respectively representing the transmitting power of the unmanned plane j and the channel gain between the unmanned plane j and the user k; the third term is the out-of-cluster interference from geosynchronous orbit satellites, p ^、 _s And g ^、 _sk The transmit power of the satellite and its channel gain with user k are shown, respectively. Since the satellite channel gain is different from the channel gain of terrestrial communications, special symbols are added to distinguish them; fourth term sigma ² Is additive white gaussian noise.

Correspondingly, if the user k is associated with UAVu, the binary variable xi _uk The coefficient is 1, otherwise 0. Specifically, the user achievable rate served by the UAV is defined by R _ku ^u The specific expression is as follows:

R _ku ^u =ξ _uk z ^～ _uk log ₂ （1+（p ^～ _u g ^～ _uk /（∑ _i=1 ^M p _i g _ik +∑ _j=1，j≠u ^U p ^～ _j g ^～ _jk +p ^、 _s g ^、 _sk +σ ² ）））

wherein z is ^～ _uk Representing downstream bandwidth resources, p, when user k is associated with the u-th UAV ^～ _u Is the transmit power of unmanned aerial vehicle u, g ^～ _uk Is the channel gain between drone u and user k.

Correspondingly, if user k is served by satellite, binary variable ζ _sk The coefficient is 1, otherwise 0. In particular, user accessibility by satellite servicesThe rate is defined by R _ks ^s The specific expression is as follows:

R _ks ^s =ξ _sk z ^、 _sk log ₂ （1+（p ^、 _s g ^、 _sk /σ ² ））

wherein z is ^、 _sk Representing the downstream bandwidth resources for user k when associated with satellite s, p ^、 _s Is the transmission power of satellite s g ^、 _sk Is the channel gain between satellite s and user k.

Step three, taking the maximized user reachable rate as an optimization target, introducing constraint conditions of power and bandwidth, and establishing an optimization problem of a downlink communication link;

the actual achievable rate of the user is defined as follows according to the association condition:

R _k =∑ _m=1 ^M ξ _mk R _km ^c +∑ _u=1 ^U ξ _uk R _ku ^u +ξ _sk R _ks ^s

in terms of constraint conditions, the transmission power of the UAV and the satellite cannot exceed the maximum transmission power limit of the system due to the limitation of system resources, and the sum of the corresponding maximum transmission power of the AP, the UAV and the satellite is defined as P _max ，P ^～ _max ，P ^、 _max There are the following constraints:

∑ _m=1 ^M p _m ＜P _max

∑ _u=1 ^U p ^～ _u ＜P ^～ _max

∑ _m=1 ^M p _m +∑ _u=1 ^U p ^～ _u +p ^、 _s ＜P ^、 _max

p _m ＞0，p ^～ _u ＞0，p ^、 _s ＞0

the constraints are also bandwidth limited, the maximum bandwidth used by a single AP, a single UAV and a single satellite are Z _m，max ，Z ^～ _u，max ，Z ^、 _s，max There are the following constraints:

∑ _k=1 ^K ξ _mk z _mk ＜Z _m，max

∑ _k=1 ^K ξ _uk z ^～ _uk ＜Z ^～ _u，max

∑ _k=1 ^K ξ _sk z ^、 _sk ＜Z ^、 _s，max

z _mk ＞0，z ^～ _uk ＞0，z ^、 _sk ＞0

the objective function and the constraint condition are comprehensively considered, the power control strategy and the bandwidth allocation strategy are defined as { P, Z }, and the finally obtained optimization problem is as follows:

max _｛P,Z｝ ∑ _k=1 ^K R _k

∑ _M∪U∪S ξ _mk +ξ _uk +ξ _sk ≤1

ξ _mk 、ξ _uk 、ξ _sk ∈｛0,1｝

and step four, based on a genetic algorithm, judging whether an optimization function is met or not by carrying out mutation and inheritance on an initialization optimization result according to mutation cross selection, selecting an optimal individual as an optimal solution of the genetic algorithm, and completing the distribution of power and bandwidth resources. The method comprises the following steps:

First, the encoding design is performed, and in the optimization problem, a binary string or an integer string may be used to represent the parameter value. For example, binary strings may be used to encode the range of values for the power control policy and the bandwidth allocation policy; initializing a population, randomly generating an initial solution, and meeting constraint conditions; then, carrying out fitness evaluation, wherein in the optimization problem, a fitness function is defined as an objective function value; then selecting operation, sorting solutions according to a target value function based on a fitness function, selecting a plurality of solutions with top ranks, then crossing and mutating, rechecking and adjusting individual parameter values after crossing operation under a binary coding strategy, or randomly changing individual codes, wherein the crossing and mutating operation also needs to ensure that the individual meets power and bandwidth constraint conditions; then iterating repeatedly, repeatedly carrying out the operations of selection, intersection and variation until reaching the condition of stopping the iteration, judging whether the iteration standard is met or not, wherein the iteration times are 1000 times, and if the iteration standard is met, entering the next step; after the optimal individual is obtained, the optimal individual needs to be decoded into actual parameter values, and for the optimization problem of power and bandwidth, the decoding operation converts binary strings or integer strings of the optimal individual into actual power and bandwidth values; and finally, analyzing and post-processing the optimized result, verifying whether the constraint condition is met, and adjusting according to the actual situation.

In another embodiment of the present application, steps S2 to S4 may further be: when a downlink communication transmission model is constructed, S1, constructing an unmanned aerial vehicle satellite communication system according to parameters such as the number, the position, the speed, the requirement and the like of the unmanned aerial vehicle and the terminal

A topology of a system, wherein a drone, a terminal and a satellite are used as nodes of the graph, a communication link between the drone and the satellite is used as an edge of the graph, and the capacity of the communication link is used as a weight of the edge.

S2, determining a source node and a sink node in the graph according to service requirements of the unmanned aerial vehicle satellite communication system, wherein for example, if the unmanned aerial vehicle needs to transmit data to a terminal, the unmanned aerial vehicle can serve as the source node, and the terminal can serve as the sink node.

And S3, constructing a network topology diagram of the unmanned aerial vehicle satellite communication system. According to the information such as the position of the unmanned aerial vehicle, the orbit parameter of the satellite, the position of the ground station and the like, the connection relation among the unmanned aerial vehicle, the satellite and the ground station, and the parameters such as the channel capacity, the transmission delay and the like of each connection are determined, and a weighted directed graph is constructed to represent the network topology structure of the unmanned aerial vehicle satellite communication system. And calculating the network capacity of the unmanned aerial vehicle satellite communication system. According to the network topology, the network capacity of the unmanned aerial vehicle satellite communication system, namely the maximum data transmission rate from a source node (ground station) to a sink node (unmanned aerial vehicle), is solved by using the maximum flow minimum cut theorem. The network capacity of the unmanned aerial vehicle satellite communication system is optimized. According to the calculation result of the network capacity, main factors affecting the network capacity, such as the position of the unmanned aerial vehicle, the orbit parameters of the satellite, the position of the ground station and the like, are analyzed, and a heuristic algorithm or a meta-heuristic algorithm is adopted to optimize and adjust the factors so as to improve the network capacity of the satellite communication system of the unmanned aerial vehicle.

And S4, establishing a Markov decision process model of the unmanned aerial vehicle satellite communication system. Modeling the network capacity optimization problem of the unmanned aerial vehicle satellite communication system as a Markov decision process, wherein a state space is information such as positions, speeds and the like of the unmanned aerial vehicle, satellites and ground stations, an action space is control such as movement, steering and the like of the unmanned aerial vehicle, the satellites and the ground stations, and a reward function is an increase or decrease of the network capacity. According to the Markov decision process model, a deep reinforcement learning algorithm suitable for the unmanned aerial vehicle satellite communication system, such as a deep Q network, a depth deterministic strategy gradient, a deep actor critique and the like, is designed, a neural network is utilized to approximate a value function or a strategy function, and learning and control of the network capacity optimization problem are realized. And according to the output of the deep reinforcement learning algorithm, evaluating the change condition of the network capacity of the unmanned aerial vehicle satellite communication system, performance indexes such as convergence, stability and robustness of the algorithm, and the like, comparing and analyzing with other algorithms, and verifying the effectiveness and superiority of the deep reinforcement learning algorithm.

The preferred embodiments of the present invention have been described in detail above, but the present invention is not limited to the specific details of the above embodiments, and various equivalent changes can be made to the technical solution of the present invention within the scope of the technical concept of the present invention, and all the equivalent changes belong to the protection scope of the present invention.

Claims

1. The method for optimizing the honeycomb-removed large-scale MIMO resource based on unmanned aerial vehicle-satellite cooperation is characterized by comprising the following steps of:

2. The method for optimizing the large-scale MIMO resources for cellular removal based on unmanned aerial vehicle-satellite cooperation according to claim 1, wherein the step S1 is specifically:

3. The method for optimizing the large-scale MIMO resources for cellular removal based on unmanned aerial vehicle-satellite cooperation according to claim 2, wherein step S2 is specifically:

4. The method for optimizing the large-scale MIMO resources for cellular removal based on unmanned aerial vehicle-satellite cooperation according to claim 3, wherein step S3 is specifically:

5. The method for optimizing the large-scale MIMO resources for cellular removal based on unmanned aerial vehicle-satellite cooperation according to claim 4, wherein step S4 is specifically:

6. The method for optimizing a large-scale MIMO resource for cellular removal based on unmanned aerial vehicle-satellite cooperation according to claim 5, wherein maximizing the user minimum rate is specifically:

x _ik =∑ _k∈Ki sqrt（p _mk ）g _mk d _k

y _k =∑ _i ^M∪U ξ _ik g _ik x _ik +ω _k

the user achievable rates for the services provided by the terrestrial APs are:

wherein sigma ² For the additive white gaussian noise power,

the overall reception rate of the user is:

R _k = R _k ^cf + R _k ^u-s

wherein R is _k ^u-s The rate of user k servicing the unmanned satellite.

7. The method for optimizing the large-scale MIMO resources based on the unmanned aerial vehicle-satellite cooperation according to claim 5, wherein the constraint condition of maximum transmitting power is specifically:

∑ _k=1 ^K ξ _mk ∑ _m=1 ^M P _mk ＜P _max

∑ _k=1 ^K ξ _uk ρ _uk ＜ρ _max

8. The method for optimizing the large-scale MIMO resources for cellular removal based on the unmanned aerial vehicle-satellite cooperation according to claim 5, wherein the problem of non-convex optimization of the large-scale MIMO for cellular removal based on the unmanned aerial vehicle-satellite cooperation is specifically:

max _{η，（x，y）Шk} minR _k =R _k ^cf +R _k ^u-s

C ₁ :∑ _M∪U ξ _mk +ξ _uk ≤1

C ₂ : ξ _mk 、ξ _uk ∈｛0,1｝

C ₃ :‖q _j -q _u ‖≤S _af ,u≠j

s.t. C ₄ ：∑ _k=1 ^K ξ _mk p _mk ＜P _max

C ₅ : ∑ _k=1 ^K ξ _uk ρ _uk ＜ρ _max

C ₆ :∑ _u=1 ^U ∑ _k=1 ^K ξ _uk ρ _sk g _uk ≤ζ _t

wherein ζ _t Interference threshold, q, tolerated for users served by satellite _u Is the horizontal position of the unmanned aerial vehicle, ρ _sk Transmitting power g for serving a satellite to a user _uk The class represents arbitrary for the channel gain between the drone and the user.

9. The method for optimizing large-scale MIMO resources for cellular removal based on unmanned aerial vehicle-satellite cooperation according to claim 5, wherein step S33 is specifically:

10. The method for optimizing the large-scale MIMO resources by removing cells based on unmanned aerial vehicle-satellite cooperation according to claim 1, further comprising solving a non-convex optimization problem by adopting a genetic algorithm to obtain an optimal solution, specifically: