Sub-band Robust GNSS Signal Processing for Jamming Mitigation

Abstract

Jammers, illegal devices, which broadcast powerful signals in the Global Navigation Satellite System (GNSS) frequency bands, can significantly degrade the performance of GNSS receivers. A possible approach to mitigate the impact of jamming is the implementation of Pulse Blanking (PB) that is an effective mitigation technique for pulsed interference. Clean samples are however required for receiver processing. In order to achieve this condition, filtering can be used: the jamming signal usually sweeps large frequency bands. When filtering is applied, the jamming signal periodically enters and exits the filter bandwidth resulting in pulsed interference. For wide-band GNSS signals, a Filter Bank (FB) can be adopted: the input signal is split into several sub-bands and PB is applied independently on each sub-band. The jamming signal impacts only one sub-band at the time resulting in a sequence of pulses in the different sub-band channels. In this paper, sub-band signal processing is combined with robust Zero-Memory Non-Linearities (ZMNLs) and a general scheme for interference mitigation is proposed. In particular, PB is replaced by ZMNLs that introduce robustness to the processing of GNSS signals. Sub-band decomposition is performed using a uniform Discrete Fourier Transform (DFT) FB, which is efficiently implemented using polyphase decomposition. Different FB configurations and different ZMNLs are considered. Hardware simulations and real GNSS data are used to demonstrate the effectiveness of the mitigation schemes proposed.