ESTIMATION OF THE CHANNEL TRANSFER FUNCTION USING HIGHER ORDER STATISTICS (HOS)
|Signal processing with HOS has not yet become a full-blown subject, but it does carry a substantial future in the creation of advanced communication systems – sonar, radar, speech, geophysical and image processing. There is a wide variety of operations that can be done with the help of HOS and particularly with higher order spectra.
The goal, the estimation of the channel transfer function, is accomplished by a simulation program that generates a non-Gaussian signal and passed through a channel which is corrupted with a Gaussian noise. The original signal is reconstructed from the noise corrupted data. This process has yielded satisfactory results.
If non-Gaussianity, non-minimum phase, or nonlinearities are the problems to be solved in different applications, then higher-order statistics offers a new set of tools, methodology and algorithms for handling some or all of these effects.