High precision technique for PPS estimation in impulsive noise environment

A technique for high precision parametric estimation of high-order polynomial phase signals (PPS) is introduced that can be used in impulsive noise environments. It is inspired by the recently proposed quasi maximum likelihood (QML) estimator with several modifications in coarse and fine estimation stages as well as in the optimization criterion. The developed technique achieves improvement in the mean squared error (MSE) for more than 20 dB with respect to the standard QML for high-order PPSs of short length and up to 20% of impulses.