Estimation of Compound K-distribution Modeling Parameters of Sea Clutter with Unknown Thermal Noise Power

摘要

This paper explores a new and efficient method for estimating the parameters of the compound K-distribution model of sea clutter when the thermal noise power is also unknown. Current methods for the model parameter estimation include the intensity moments approach that assumes knowledge of the thermal noise power and a three-dimensional (3-D) curve fitting approach that is very computationally intensive. The proposed method integrates the intensity moments with a nonlinear one-dimensional (1-D) curve-fitting procedure to also allow for an efficient estimate of the thermal noise power. It is also extended to incorporate fractional moments used under the condition of known clutter-to-noise ratios. The performance of the new approach is demonstrated using both simulated and real sea clutter observations.