【无人机】基于粒子群优化干扰受限下无人机群辅助网络附matlab代码

本文链接：https://blog.youkuaiyun.com/matlab_dingdang/article/details/128006041

✅作者简介：热爱科研的Matlab仿真开发者，修心和技术同步精进，matlab项目合作可私信。

🍎个人主页：Matlab科研工作室

🍊个人信条：格物致知。

更多Matlab仿真内容点击👇

智能优化算法神经网络预测雷达通信无线传感器

信号处理图像处理路径规划元胞自动机无人机电力系统

⛄ 内容介绍

本文研究了三种在实际的空地（ATG）信道模型下的基本部署设计，即最小数量的无人机，它们的最佳部署位置和最佳发射功率分配。由于这三个设计目标是耦合的，结合平衡信号，提出了一种基于粒子群优化（PSO）的方案干扰加噪声比（SINR）发射功率分配。通过利用封闭形式表达式在SINR平衡最优功率分配和由此产生的SINR中，提出了基于PSO的方案迭代优化无人机的数量，然后优化它们的位置，直到每个无人机的SINR用户满足其所需的最小值。为了改进所提出方案的收敛性设计了一些方案，通过分析系统，对无人机的最小数量进行初步估计在使用K-means聚类技术初始化UAV位置之前，求和速率容量。最后开发了功率微调方案以进一步降低总发射功率，广泛的模拟以确认所提出方案的良好性能。

⛄ 部分代码

function [SINR,P ] = Flow_UAV(ITER,X,users_sets)

k_center = size(X,2)/3;

for i = 1: k_center

centerpoint(i,:) = [X(i*3-2),X(i*3-1)];

h(i) = X(i*3);

end

N_users = users_sets;

u_distance = 500;

%all_users = u_distance * rand(200,2);

%load('alluser.mat', 'all_users');

s=ITER;

rng(s);

all_users = 500*rand(users_sets,2);

%% pathloss max choice

N_total = 20;

placement = zeros(N_total,N_users);

for u = 1:N_users

dist_u = zeros(k_center,1);

for N = 1:k_center

dist_u(N) = norm(all_users(u,:) - centerpoint(N,:));

pathloss(u,N) = pathloss3D(dist_u(N),h(N));

end

[N_dist, N_belong] = min(pathloss(u,:));

label(u) = N_belong;

n=hist(label,[1:40]);

for N_tot = 1:N_total

if n(N_tot)>20

pathloss(u,N_tot)=[10^30];pathloss(u+1,N_tot)=[10^30];

end

end

[N_dist, N_belong] = min(pathloss(u,:));

label(u) = N_belong;

placement(label(u),u) = 1;

dist_all(u) = N_dist;

power_normolization(1,u)=1;

yita(u) = 10^-12;

end

%% pathloss

alpha = [0.0001,0.001];

for u_row = 1:N_users

for u_col = 1:N_users

R_deno = norm(all_users(u_row,:)-centerpoint(label(u_row),:));

R_nume = norm(all_users(u_row,:)-centerpoint(label(u_col),:));% the distance between the u_col-th user and the u_row-th user's BS.

g(u_row, u_col,:) = 1/pathloss3D(R_nume, h(label(u_col)));

end

r(u_row) = 1/pathloss3D(R_deno, h(label(u_row)));

end

diag_r = diag(r);

f = g-diag_r*eye(N_users);

for K = 1:k_center %damping factors

for i = 1:N_users

for j = 1:N_users

A(i,j) = alpha((label(i)==label(j))+1);

end

end

DC(:,:,K) = diag(1./r) * ( A.*f+ 10*yita'*placement(K,:) ); %SINR model

[a(:,:,K),b(:,:,K)] = eig(DC(:,:,K));

eigenvalue = diag(b(:,:,K));

lamda(K) = max(eigenvalue);

for i = 1:length(DC(:,:,K))

if lamda(K) == eigenvalue(i)

break;

end

end

y_lamda(:,K) = a(:,i,K);

end

[rho,index] = max(lamda);

SINR = 1/rho;

P(:) = 0.1*y_lamda(:,index)/(placement(index,:)*y_lamda(:,index)); %real power

end

⛄ 运行结果

⛄ 参考文献

[1] Liu W , Niu G , Cao Q , et al. Particle Swarm Optimization for Interference-Limited Unmanned Aerial Vehicle-Assisted Networks[J]. IEEE Access, 2020, 8:174342-174352.

⛄ Matlab代码关注

❤️部分理论引用网络文献，若有侵权联系博主删除

❤️ 关注我领取海量matlab电子书和数学建模资料