Efficient approximation for linear and non-linear signal representation

Abstract

This paper focuses on optimum representation for both linear and non-linear type signals which have a wide range of applications in the analysis and processing of real-world signals, that is, noise, filtering, audio, image etc. Accurate representation of signals, usually is not an easy process. The optimum representation is achieved by introducing exponential bases within multiplicative calculus which enables direct processing to reveal the unknown fitting parameters. Simulation tests confirm that the newly introduced models produce accurate results while using substantially less computation and provide support for applying the new model in the field of parametric linear, non-linear signal representation for processing.