ML DA CA symbol synchronization using framemwise constant channel gain estimates for nonselective fading channels

Investigates a maximum-likelihood (ML) decision/data-aided (DA) channel-aided (CA) symbol synchronization algorithm operating on a noisy M-PSK signal transmitted over a slowly time-varying nonselective Rician fading channel. Each of the channel gain estimates used by the DA CA algorithm is held constant over a frame of F symbols. Compared to the hypothetical case where perfect channel gain estimates are available, the tracking performance of this algorithm is degraded by these imperfect channel gain estimates. The authors investigate how the linearized timing error variance depends on the channel characteristics (carrier-to-multipath ratio C/M, fading bandwidth) and on the frame length F. They show that the value of F that minimizes the tracking error variance, decreases with increasing fading bandwidth and decreasing C/M