Ricerc@Sapienza

Time Reversal (TR) is a prefiltering scheme mostly analyzed in the context of centralized and synchronous IR-UWB networks, in order to leverage the trade-off between communication performance and device complexity, in particular in presence of multiuser interference. Several strong assumptions have been typically adopted in the analysis of TR, such as the absence of Inter-Symbol / Inter-Frame Interference (ISI/IFI) and multipath dispersion due to complex signal propagation. This work has the main goal of comparing the performance of TR-based systems with traditional non-prefiltered schemes, in the novel context of a distributed and uncoordinated IR-UWB network, under more realistic assumptions including the presence of ISI/IFI and multipath dispersion. Results show that, lack of power control and imperfect channel knowledge affect the performance of both non-prefiltered and TR systems; in these conditions, TR prefiltering still guarantees a performance improvement in sparse/low-loaded and overloaded network scenarios, while the opposite is true for less extreme scenarios, calling for the developement of an adaptive scheme that enables/disables TR prefiltering depending on network conditions.