Wavelet-Based Subspace Method for Blind Estimation of Time-Varying Channels

A novel linear algorithm is proposed in this paper for estimating time-varying channels only relying on second order statistics. The proposed methods are applicable to channels where time-varying impulse response can be described as a linear combination of a finite number of wavelet basis functions. It is shown that the estimation of the wavelet expansion coefficients is equivalent to estimating the second-order parameters of an unobservable single input many output process, which are directly computed from the observation data. By exploiting this equivalence, blind subspace methods are applicable. Some illustrative simulations are presented.