Signal recovery and amplitude-phase tracking in wireless communications

This paper addresses the problem of demodulating a digital communications signal in the presence of unknown channel complex gain. We derive the Cramer-Rao lower bound for tracking the channel gain in the presence of additive white Gaussian noise. This bound establishes effective lower bounds for the signal-to-noise ratio on the channel, given the statistics of the channel gain variations. Several approaches for tracking the gain are derived based on the application of Bayes theorem, and the approximation of soft decision symbol statistics by a Gaussian density. The performance of the tracker is examined using simulated and real data.