A Novel Adaptive Architecture: Joint Multi-targets Detection and Clutter Classification

A novel adaptive architecture is conceived to face with the problem of multiple point-like targets detection buried in Gaussian disturbance with the lacking of targets information, including their positions, number, and angles of arrival. To this end, a target-rich scenario where the clutter properties vary over the range profile is considered. Such distinct clutter properties are modeled in terms of different interference covariance matrices that provide the basis to jointly classify clutter and targets over the range. Specifically, suitable estimates of the unknown parameters are figured out by adopting the expectation-maximization algorithm together with a grid search approach. Then, a decision scheme based upon the Likelihood Ratio Test is exploited along with the estimated values. The performance assessment, conducted resorting to simulated data, highlights the effectiveness of the proposed scheme in heterogeneous interference.