Maximum likelihood multichannel estimation under reduced rank constraint

This paper deals with the maximum likelihood estimation of the multichannel impulse response in a mobile communication system whose base stations are equipped with antennas arrays. The following problem is solved: using the training sequence, find the maximum likelihood multichannel impulse response from one mobile to the base station under a reduced rank constraint in the presence of Gaussian noise and jammers with unknown covariance matrix. Our results find applications in equalization (the reduced rank channel estimate can be used in a Viterbi algorithm), and in the estimation of the directions of arrival (DOA) of the paths from the mobile to the base station. In this last application, a MUSIC like algorithm is developed using the estimated channel subspace.