Radio Frequency Interference Mitigation via Nonnegative Matrix Factorization for GNSS

A radio frequency interference (RFI) mitigation approach based on nonnegative matrix factorization (NMF) for global navigation satellite systems (GNSS) signals is proposed. The proposed approach employs NMF to separate the interference from the GNSS signals, and it can be deployed in a supervised or semiblind manner. The supervised NMF framework assumes prior knowledge about the RFI, whereas its semiblind counterpart does not require any
a priori
information about the RFI. Results indicate that both schemes are able to mitigate narrow and wideband RFI signals, outperforming Kalman, notch filter, and wavelet-based techniques, enabling GNSS signal acquisition even in scenarios where the interference is 50 dB stronger than the GNSS signals. In addition, the proposed approach is able to mitigate multiple different types of RFI corrupting the received GNSS signal.