Narrowband interference suppression in a spread-spectrum system using vector space methods

The author develops vector-space interference-excision (VSIE) methods that suppress the sidelobes of a sine-wave interferer, in addition to the central lobe, while removing little signal power. THE (threshold excision) and VSIE are applied to synthetic data sets consisting of a binary PN (pseudonoise) sequence in Gaussian background noise and sine-wave interference. Signal-to-noise ratios (SNRs) and bit-error rates are given for four models in which up to 40% of the spectrum exceeds a threshold 7 dB above the average background level. The SNR after VSIE is close to the SNR with no interference present and as much as 8 dB better than the SNR after THE.<
>