The effective coherence time of common channel models

Channel coherence time is commonly used to characterize the rate at which the channel changes. The recently introduced effective coherence time measures the channel change rate from an estimation perspective. It was shown that the effective coherence time is related to the capacity of the non-coherent channel, in which the receiver has no prior knowledge of the channel, and tries to estimate it from the received signal. In this paper we present novel bounds on the non-coherent channel capacity with a peak power constraint or a constraint on the second and fourth moments of the transmitted signal. This bounds are characterized only by the system signal-to-noise ratio and the effective coherence time. We then focus on the characteristics of the effective coherence time. We present two alternative formulas for the effective coherence time, and give closed form expressions for the effective coherence time of the auto regressive and the Clarke's channel models.