SNR estimation and jamming detection techniques using wavelets

Abstract

An SNR estimation approach and a jamming detector based on wavelet transform theory are presented. The SNR estimator is an in-service, non-data-aided estimator that operates on M-PSK and QAM modulated signals transmitted over baseband CWGN channels. The signal and noise power are separated through a non-linear wavelet technique known as denoising. Two wavelet-based estimators are presented. The first method uses hard-thresholding which extracts the amplitude trend over one or several symbol periods, depending on whether the modulation is constant or multi-level envelope. The second method uses adaptive soft-thresholding and applies a self-similarity criterion between the signal and wavelet. A SNR Moments estimator was also developed as a reference for evaluation purposes. A jamming detector based on discontinuity recognition using wavelets is presented. The detector is implemented for constant-envelope modulation schemes, leaving the multi-level case for future research.