Analysis of notch filtering methods for narrowband interference mitigation

Interference, intentional and unintentional, is becoming a major threat for today's GNSS receivers. Several authors have presented notch filtering as an efficient mean to mitigate continuous wave or narrowband interference. Most of those studies, however, have focused only on the amount of signal energy preserved after filtering and limited amount of research has been done in analyzing the distortion of the autocorrelation function caused by the filtering process and its impact on the tracking performance. This study aims to analyze the effect of different notch filters on the autocorrelation function distortion. Research is conducted using simulations with actual PRN codes and for different modulations, namely BPSK(1), BPSK(10), BOC(1,1), and CBOC(6,1). Software simulations are then validated in a GNSS hardware prototype and it is shown that the results from the hardware prototype are consistent with the simulation results. Code and carrier tracking biases and error variance are defined as the key tracking performance indicators. Analysis of the notch filtering methods show that some of the proposed filter structures for filtering interference in GNSS receivers do not work well in practice and a great care should be taken when implementing filtering in the GNSS signal processing chain. In some cases, notch filtering should not be applied even though there is an interference signal present, as it will only degrade the GNSS receiver performance.