Dear Researchers,
I have a model in which I extracted Signal Subspace (SS),Qs, from the measured Covariance matrix. Now I wan to extract the noise subspace (of course without EVD decomposition method). my question is: If I project the null space of Qs as follow:
Projected_Qs=eye(M) - Qs x inv(Qs' x Qs) x Qs'
Is that mean I get noise subspace ,Qn, please? in other word, is Projected_Qs=Qn?
I appreciate your answers.