Research On Interference Mitigation For Compressed Sensing Based IR-UWB Systems

Impulse radio ultra-wideband (IR-UWB) is a new type of wirelesscommunication system that satisfies the demand for higher data rates at short ranges.Due to its prominent advantages, such as large system capacity, small energyconsumption, strong penetration ability, and low probability of interception,IR-UWB has attracted widespread interest and intensive study. However, advantagesare accompanied by challenges. Using an ultra-short duration pulse, typically atnanosecond or sub-nanosecond levels in the time domain to carry information,IR-UWB occupies extremely high bandwidth in the frequency domain. And，highspeed analog-to-digital converter is hardly achieved with low cost.Compressed sensing (CS) is an emerging theory that enables the reconstructionof sparse signals from a small set of random measurements. Because of the sparsityof IR-UWB signals in the time domain, CS makes it possible to operate atsub-Nyquist rates for IR-UWB communications where Nyquist sampling representsa formidable challenge. However, interference signal can greatly degrade the systemperformance. The focus of this paper is to design the interference mitigationschemes using CS.The interference in IR-UWB communications mainly includes narrowbandinterference (NBI) and multi-user interference (MUI). This paper proposes thedesign of a new compressive measurement matrix that can mitigate NBI and MUI inIR-UWB communication systems under the CS framework. To enhance the existingIR-UWB receiver structure, an information feedback loop from Digital Back-end toAnalog Front-end measurement waveform generator is added. In the aspect of NBImitigation, it is sparse in the frequency domain. The subspace of NBI is estimatedthrough the orthogonal matching pursuit (OMP) restructuring algorithm. Byexploiting the subspace, the null space of NBI is achieved. A new measurementmatrix is structured based on the null space information fed back to themeasurement waveform generator. For the MUI mitigation, it is sparse in the timedomain. We creatively put forward the concepts of reference position pulsewaveform and Time-Hoping matrix with the information of TH sequence. Afterreceiving the reference waveform, its subspace is estimated and fed back. Thesubspace is also used to structure the measurement matrix. Under the CS-UWBframework, NBI and MUI signal can be mitigated or cancelled at a sampling rate farbelow the Nyquist rate when projecting the received UWB signal onto themeasurement matrix without loss of useful signal energy. Furthermore, consideringthe research on interference mitigation in the time domain and the frequency domain, this paper proposes the combined mitigation of NBI and MUI, and discusses thesystem demodulation scheme aiming at the interference mitigation based on thesystem characteristics, which make the research perfect and more systematic.