Search

A new technique is developed for intellectual property protection of virtual cores that initiate digital signal processing algorithms.

Mathematical bounds that are independent of array locations establish the influence of temporal and spatial sampling on the variance of the frequency estimates of superimposed narrowband signals. An increase in the fourth central moment enhances the resolution ability of the array. The resolution is independent of the array aperture. The product of the number of sensors and the second central moment characterizes the estimator variance lower bound.

Frequency estimation of a complex sinusoid corrupted by complex-valued multiplicative noise is discussed.

Deals with a study which presented a framework for deriving multiscale and hybrid functionally expanded approximators based on the reproducing kernel Hilbert spaces. General framework for nonlinear mapping functional expansions; Applications of the framework; Numerical results.