cvx_begin quiet
variable g0(N,1)
minimize sum(abs(g0(1:N)))
subject to
c*g0(1:N)==m;
for i=1:Ka
norm([Tw(2*i-1,:);Tw(2*i,:)]*g0(1:N))<=0.05;
end
for i=1:L
D(i,:)*g0(1:N)==0;
end
cvx_end
where, c,Tw,D are known
cvx_begin quiet
variable g0(N,1)
minimize sum(abs(g0(1:N)))
subject to
c*g0(1:N)==m;
for i=1:Ka
norm([Tw(2*i-1,:);Tw(2*i,:)]*g0(1:N))<=0.05;
end
for i=1:L
D(i,:)*g0(1:N)==0;
end
cvx_end
where, c,Tw,D are known
First of all: please do not post multiple questions on the same topic. I have deleted your other post.
CVX is telling you the model is infeasible. I’m inclined to trust it. Consider trying other solvers to see if they produce the same results.
sedumi and sdpt3 are all tried. are there other solvers ?
Sure, MOSEK (if you have the Professional license) and ECOS (you have to install it yourself). But honestly, the solver is almost certainly not the problem. Your model is simply infeasible.