Phase-Based Carrier Frequency Estimators for Linear Modulations over Selective Fading Channels

Phase-based frequency estimators over frequency-selective fading channels are examined. The implementation complexity of phase-based estimators is low due to their explicit structures which require no numerical search. The statistical information of the measurement phase noise (MPN) is crucial for phase-based estimators. The exact probability density function of the MPN over frequency-selective fading channels is first derived. Then, two approximations are proposed to further reduce the computational complexity of the estimators. In our new phase models, the magnitudes of received signal samples enter in determining the covariance of the MPN, and give adaptive weights to the phase measurements. In contrast, only phase information is exploited in the traditional phase-based frequency estimators due to the approximate MPN model used. Our simulation results demonstrate that better performance can be achieved by using our improved MPN models over frequency selective fading channels. Finally, the effects of the periodic length of the training sequence on the performance of frequency estimation are studied.