New Thresholding Technique in DCT Domain for Interference Mitigation in GNSS Receivers

Abstract

The presented method is a DCT mitigation thresholding technique (DCT-MTT) for narrowband interference reduction in Global Navigation Satellite System (GNSS) receivers. First, the received signal immersed in an Additive White Gaussian Noise (AWGN), is multiplied, in time domain, (sample by sample) by a Tukey window of the same length. Then, the DCT transform is applied. Next, the transformed signal is divided in nonoverlapped packets. Each one can be viewed as a non-interfered packet (if it has roughly the same variance as the estimated unknown variance of the AWGN) that must be conserved, or considered as an interference packet (if its variance is significantly greater than the variance of the AWGN) that should be thresholded. The conservation (or inversely the thresholding) of a packet is achieved by the use of DONOHO's Universal-Threshold apart from that the variance is estimated based on the statistical sampling theory. The Final step consists in the application of the inverse DCT to obtain a good approximation of the received interference-less signal. The results obtained from several simulations confirm that the suggested strategy outperforms, in term of signal quality restoration, the conventional methods.

Envelope Detection by Shannon Energy Calculation in DCT Domain and DFS-Based Notch Filter for Interference Mitigation in GNSS Receivers

“…PSDs of the signal before and after the interference reduction [5,7,18]. The ambiguity function S(τ,Fd) which results from The satellite acquisition operation [6,13,18]. The correlation coefficient showing how is the degree of similitude between the retrieved signal and the original clean signal [13].…”

“…It is based on an N-sigma excision algorithm based on FFT implied to eliminate the narrow-band interferences, without removing other frequency components of the useful signal. Additionally, Khezzar et al [13] proposed a new thresholding technique applying the well-known Donoho's algorithm in the DCT frequency domain. However, the estimation of the variance is achieved based on the statistical sampling theory.…”