A phase estimator for complex signals using evolutionary algorithms

This paper presents a new method of phase estimation that is based on evolutionary algorithms. Unlike traditional phase estimators, this method does not require detailed a priori knowledge of the signal. The estimator operates by generating a model of the signal and optimising the model parameters including phase, to provide the best match in the frequency domain. The results show that the estimation process is capable of providing a good phase estimate of the signal, even in significant noise and for a wide range of waveforms.