Mean and covariance based estimation of multiple access UWB channels

Traditional pulse position modulation based ultra-wideband (UWB) communication system transmits short pulses whose positions carry user's data information. Using different time-hopping sequences, multiple access is enabled. We address blind multiuser channel estimation after following a pulse-rate discrete-time channel model. The model can be regarded in a tri-linear structure and also resembles a code-division multiple access (CDMA) system with newly defined hopping-code dependent matrices for each user. Considering that either the mean or covariance of received signals contains sufficient information for all unknown channels, least squares and covariance matching ideas are successfully applied for channel estimation. Performance of each estimator is analyzed and also verified by computer simulations.