Adaptive reduced-rank interference suppression for DS-UWB systems based on switched approximation of basis functions

A novel low-complexity reduced-rank adaptive interference suppression scheme is proposed in this work. The new scheme is based on a switched approximation of basis function and we consider its application to a multiuser direct sequence spread-spectrum ultra-wideband (DS-UWB) system. In the proposed scheme, a projection matrix is chosen instantaneously from a set of matrices that consists of assigned basis functions for dimensionality reduction. A least-mean squares (LMS) adaptive algorithm is then developed to estimate the coefficients of the reduced-rank adaptive filter and a complexity analysis for the adaptive algorithms is given. Simulations show that the proposed scheme has remarkable interference suppression performance with very low complexity.