Nonwhite noise/interference rejection via blind adaptive filtering

The paper deals with the issue of non-invasive filtering of the non-white narrow-band interference within the spread spectrum signals. This is done in two steps. First, the signal plus interference, as one quantity and the replica, as the separate one-are whitened in parallel due to the power spectrum of the interference. Second the obtained products are processed in accordance with the ML crosscorrelation principle. The simulation rejects the interference down to -20 dB for a signal-to-interference bandwidth ratio B/b/spl ap/10. If the interference is not known, the adaptation is engaged.