Matrix decomposition with contraints

36 Views Asked by At

I am planning to decompose the Hermitian matrix X (which is function of r, $\theta$) which is a square matrix as $X=ABA^H$.

Let us assume size of A is NxL where L is the no. of Eigen vectors which is function of (m,n). If we decompose X into L Eigen vectors and every $i^{th}$ Eigen vector A(:,i) is very arbitrary (EVD).

I would like to put a constraint on A such that; $A(:,i)=\phi_{m,n}(r)e^{im\theta}$.

I would like to calculate $\phi_{m,n}$ (L=m $\times$ n) vectors).

Kindly let me know if this doesn't make sense. Thank you very much.