Morphological Detector for Multilevel Signals in epsilon-Noise

Abstract

The novel approach was developed for multilevel signal detection in channels with impulsive non-Gaussian noise. This approach consists of using morphological nonlinear image filtration principles for two dimensional signals. It is a new method of signal demodulation, using three - dimensional image processing algorithms. Successful results of this morphological detector encourage more investigation towards using image processing theory and algorithms for two dimensional signal processing. As can be seen in the example in section IV, this new approach of reusing well developed and extensively developing image processing has significantly improved performance.