%仅考虑迭代优化计算资源
%输入为用户数，每个用户的数据量，每个用户的和速率
function [Cost_opt1,F_opt] = optimization_fij(K,I,r,B)
global ku;
global C;
global Sigma; %加权系数
global P; %用户的发射功率
global F_max; %无人机计算资源约束
%global B_max; %带宽资源约束
global ti_qos; %卸载的时间QOS约束
global Epsilon;
Epsilon = 0.01;
%cost_last = 0;
%while(1)
cvx_begin
cvx_precision best
cvx_expert true
variable F_temp(K,1)
%variable B_temp(K,1)
expression cost_opt
%cost_opt = kuCsum(I.F_temp.^2);%+SigmaC*sum(I.inv_pos(F_temp));% + Psum(I.inv_pos(B.r)) + Sigmasum(I.inv_pos(B.r));
for user=1:K
%每个用户的能耗和时间加权成本
cost_obj(user) = kuI(user)C (F_temp(user)^2) + SigmaI(user)Cinv_pos(F_temp(user));%+ PI(user)/(B_temp(user)r(user)) +…Sigma(I(user)*C/F_temp(user) + I(user)/(B_temp(user)r(user)));
end
cost_opt = sum(cost_obj)
minimize cost_opt
subject to
for user=1:K
F_temp(user) >= 0;
F_temp(user) >= I(user)Cinv_pos(2ti_qos-I(user)*inv_pos(B(user)*r(user))) …
+(I(user)*inv_pos((B_temp(user)r(user)))-2ti_qos) * F_temp(user) + I(user)*C <= 0;
end
sum(F_temp) <= F_max;
cvx_end
F_opt = F_temp;
Cost_opt1 = cost_opt;
%B_opt = B_temp;
% cost_opt = cvx_optval;
% if (cost_opt - cost_last) <= Epsilon
% Cost_opt1 = cost_opt;
% break
% end
% cost_last = cost_opt;
% end
end
UAV communication? resource allocation or trajectory optimization?