H. Ye et Rd. Degroat, MAXIMUM-LIKELIHOOD DOA ESTIMATION AND ASYMPTOTIC CRAMER-RAO BOUNDS FOR ADDITIVE UNKNOWN COLORED NOISE, IEEE transactions on signal processing, 43(4), 1995, pp. 938-949
This paper is devoted to the maximum likelihood estimation of multiple
sources in the presence of unknown noise, With the spatial noise cova
riance modeled as a function of certain unknown parameters, e.g., an a
utoregressive (AR) model, a direct and systematic way is developed to
find the exact maximum likelihood (ML) estimates of all parameters ass
ociated with the direction finding problem, including the direction-of
-arrival (DOA) angles Theta, the noise parameters alpha, the signal co
variance Phi(s), and the noise power sigma(2). We show that the estima
tes of the linear part of the parameter set Phi(s) and sigma(2) can be
separated from the nonlinear parts Theta and alpha. Thus, the estimat
es of Phi(s), and sigma(2) become explicit functions of Theta and alph
a. This results in a significant reduction in the dimensionality of th
e nonlinear optimization problem. Asymptotic analysis is performed on
the estimates of Theta and alpha, and compact formulas are obtained fo
r the Cramer-Rao bounds (CRB's), Finally, a Newton-type algorithm is d
esigned to solve the nonlinear optimization problem, and simulations s
how that the asymptotic CRB agrees well with the results from Monte Ca
rlo trials, even for small numbers of snapshots.