Performance of Bayesian estimation based variable rate variable power MQAM system

In this paper, we generalize the algorithms of our previously proposed Bayesian estimation based variable rate variable power multilevel quadrature amplitude modulation (VRVP-MQAM) system to incorporate for the first time a maximum a posteriori (MAP) channel predictor and a MQAM scheme adopted with practical constellation sizes. Based on a pilot symbol assisted modulation (PSAM) scheme, we evaluate the performance of our proposed VRVP-MQAM system over a Rayleigh flat-fading channel. We demonstrate in our simulation results that the proposed rate and power algorithms that are derived based on a Bayesian bit error rate (BER) estimation and the second order statistical characterization of the channel state information (CSI) outperforms in terms of spectral efficiency and average BER. This improvement is confirmed by comparison with an alternative rate and power algorithm which exploits an ideal CSI assumption