Modified quasi‐maximum likelihood estimator for polynomial phase signal

Abstract

In this study, the authors focus on the selection of instantaneous frequency estimates in the quasi-maximum likelihood estimator for polynomial phase signal. More specifically, they are interested in the case of short signal observations at low signal-to-noise ratios. In the first part of this work, the benefit of instantaneous frequency estimates selection for the above estimator performances is highlighted. Then the authors propose a modified quasi-maximum likelihood estimator based on a determinist instantaneous frequency estimates selection. Numerical studies provided at the end of this study showed clearly that the performances of the proposal in terms of root mean square error and signal to noise ratio threshold compared to the classical approach was better, in the considered case of short signal observation in a noisy environment.