Adaptive detection using low rank approximation to a data matrix

Using an accurate formula for the error in approximating a low rank component, we calculate the performance of adaptive detection based on reduced-rank nulling. In this principal component inverse (PCI) method, one temporarily regards the interference as a strong signal to be enhanced. The resulting estimate of the interference waveform is subtracted from the observed data, and matched filtering is used to detect signal components in the residual waveform. We also present a generalized likelihood-ratio test (GLRT) for adaptively detecting a low rank signal in the presence of low rank interference. This approach leads to a test which is closely related to the PCI method and extends the PCI method to the case where strong signal components are present in the data. A major accomplishment of the work is our calculation of the statistics of the output of the matched filter for the case in which interference cancellation and signal detection are carried out on the same observed data matrix. That is, no separate data is used for adaptation. Examples are presented using both simulated data and real, active-sonar reverberation data from the ARSRP, the Acoustic Reverberation Special Research Program of the Office of Naval Research     

Data-Adaptive Detection of a Weak Signal

Motivated by a form of the likelihood-ratio-tesf statistic for detection of a rank-one Gaussian signal in colored Gaussian noise, we apply our earlier technique for estimation of a low-rank signal to the problem of estimating and subtracting the waveform of a strong sinusoidal interference prior to detection of a weak sinusoidal signal. We consider the difficult case in which samples of data are taken over a short interval of time or space and the frequencies of the sinusoidal signal and sinusoidal interference are more closely spaced than the reciprocal of the extent of the aperture. The method can be applied to cases of nonsinusoidal and/or random signals and interference. The most important assumption is that when the samples of the interference are arranged in matrix form the matrix is approximately of low rank in the sense that, with high probability, the interference-only matrix can be well approximated by a matrix of low rank.