Closed-Form Fractional-Moment-Based Estimators for K-Distributed Clutter-Plus-Noise Parameters

In this paper, the problem to estimating the parameters of the K-distributed clutter plus noise is addressed. The existing nonasymptotic z-log-z estimators are only applicable to multiple-pulse data due to the use of the harmonic mean, which is undefined for single-pulse reverberation data. When either single- or multiple-pulse data are used, new closed-form estimators are derived in this work. They combine two interesting statistical ratios based on fractional positive- and negative-order moments so that the hypergemetric functions are absolutely eliminated. For single-pulse transmission, the proposed estimators demonstrate computational advantages compared to the existing approaches. Besides, in the multiple-pulse case, they yield about the same accuracy as the closed-form z-log-z-based estimators obtained in previous work.