Modeling and Mitigation of Narrowband Interference for Transmitted-Reference UWB Systems

Transmitted-reference ultra-wideband (TR-UWB) systems, in conjunction with autocorrelation-receivers (AcR), can collect all energy of the multipath channel response at low complexity. Unfortunately the AcR front-end also collects the energy of narrowband interference (NBI) signals, which severely degrades its performance. We present an elaborate statistical analysis of the NBI for a multichannel AcR front-end. Our results show specific correlation properties, which can be exploited to mitigate the NBI by post-processing the AcR output samples. This explains how linear signal processing algorithms can be used to efficiently suppress the cross-terms arising in the nonlinear AcR between all input signals: the UWB, NBI, and noise signals. Linear least squares (LS) and minimum mean squared error (MMSE) receivers are investigated. To support our analysis, computer simulations are conducted using an IEEE802.11a WLAN signal as NBI.