Joint maximum-likelihood detection and estimation of space-time distributed multi-path parameters

Several parametric approaches for estimating the angular spread of a signal as well as its direction have been proposed previously. These approaches require that the number of signal sources should be less than the number of antennas. Furthermore, they usually consider the angle and the angular spread only. In this paper we extend the idea by modeling the received (CDMA) signal using reasonable distribution functions in terms of time as well as space parameters. The particular distributions of the angle and delay functions chosen herein are the Gaussian distribution. Our estimation algorithm based on the maximum likelihood (ML) method estimates the angles, the variance of the angular spreads, the time delay, the variance of the delay spreads, and finally the power of the individual multipath signals. The proposed method has been successfully applied to field measurement data collected in differnet regions throughout South Korea to extract representative channel parameters.