Statistical analysis of the nonhomogeneity detector for non-Gaussian interference backgrounds

Abstract

We derive the nonhomogeneity detector (NHD) for non-Gaussian interference scenarios and present a statistical analysis of the method. The non-Gaussian interference scenario is assumed to be modeled by a spherically invariant random process (SIRP). We present two methods for selecting representative (homogeneous) training data based on our statistical analysis of the NHD for finite sample support used in covariance estimation. In particular, exact theoretical expressions for the NHD test statistic probability density function (PDF) and its moments are derived. Additionally, we note that for SIRP interference, a simple transformation of the NHD test statistic admits an elegant representation as the ratio of a central-F distributed random variable and a beta distributed loss factor random variable. Performance analysis of the NHD is presented using both simulated data and measured data from the MCARM program.