Capacity measure for finite state Markov modeling of the phase process in flat fading channels

We investigate finite state Markov modeling of the phase process in flat fading channels. Phase modulations are often used in mobile communication systems. They rely on the fact that the receiver can estimate the channel phase with acceptable accuracy. In the absence of pilot signals or when the channel varies frequently, this becomes a very challenging task. Inaccuracy in phase estimation results in dramatic performance degradation. We propose mapping of the channel phase process into a finite state Markov model. In particular, we study the effect of the number of phase quantization levels and thresholds on the capacity under different signal-to-noise ratio and fading correlation conditions. In BPSK signaling, non-uniform 4-state phase quantization results in noticeable capacity enhancement under all signal-to-noise and fading correlation conditions, while in low signal-to-noise conditions, the 8-state channel outperforms the 4-state channel slightly. Both constant envelope and Rayleigh envelope models are considered in the numerical capacity analyses. The capacity improvements are also compared with the improvements in finite state Markov mapping of the fading envelope.