Maximum likelihood DOA estimation in cyclic cumulant domains

Based on the fact that a finite-sample cumulant estimation must have a noise effect, combining the cyclic cumulant and maximum likelihood algorithms (ML), a novel DOA estimation algorithm, the maximum likelihood DOA estimation in the cyclic cumulant domain (MLCCD), is proposed. This algorithm can be used not only in white noise and uncoherent signals fields, but also in correlated noise and coherent signals fields. The statistical properties of the algorithms are examined. Simulations are presented to study the properties of MLCCD estimators. The proposed algorithm is also shown to perform better than the classical ML algorithm. This algorithm combines cumulant and ML successfully, which allows application of cumulants to array processing