Low-Complexity Subspace Methods for Channel Estimation and Synchronization in Ultra-Wideband Systems

We consider the problem of low-complexity channel estimation in digital ultra-wideband receivers. We extend some of our recent sampling results for certain classes of parametric non-bandlimited signals and develop several methods that take advantage of transform techniques to estimate channel parameters from a low-dimensional subspace of a received signal, that is, by sampling the signal below the Nyquist rate. By lowering the sampling rate we reduce computational requirements compared to current digital solutions, allow for slower A/D converters and potentially significantly reduce power consumption of digital receivers. Our approach is particularly suitable for indoor wireless sensor networks, where low rates and low power consumption are required. One application of our framework to high-resolution acquisition in ultra-wideband localizers is also presented.