Improved maximum likelihood spectrum for direction of arrival (DOA) estimation

A variation of J. Capon's (1969) maximum-likelihood (ML) spectral estimation algorithm is described for calculating direction of arrival using a short linear array. The proposed modification involves altering the computed covariance spectral matrix by setting its lowest eigen value to a small, positive constant λ'(min). Theoretical analysis indicates that as λ'(min) is reduced, the improved ML spectrum tends towards the Pisarenko estimate (1972), thereby attaining a higher resolution than Capon's original ML technique while at the same time retaining its robustness. Simulation results for an eight-element array are provided to illustrate the performance of the IML spectrum in resolving two incoherent sources in the presence of background noise