flying
Posts:
10
Registered:
4/13/12


Re: How can i figure this 'nonlinear least square problem'?
Posted:
Jun 20, 2012 8:04 AM


Torsten <Torsten.Hennig@umsicht.fraunhofer.de> wrote in message <e17f48745e0244569154ed7db21536b1@v9g2000yqm.googlegroups.com>... > I think because A is complex, A = A1 + i*A2, the problem will not be > linear in A. > Note that multiplying out > (A*M1*exp(j*theta*n)M2)*conj( A*M1*exp(j*theta*n)M2) > and differentiating the term A*M1*conj(A*M1) with respect to A1, e.g., > gives (2*A1+A2^2)*M1*conj(M1), > which is linear in A1, but not in A2 (reversed if you differentiate > with respect to A2). > > Best wishes > Torsten.
Thanks.

