Bit Error Floor of 16-QAM in the Presence Of Phase Nosie

This paper presents theoretical analysis on deriving the bit error floor (BEF) of Gray coded 16-ary quadrature amplitude modulation (16-QAM) in the presence of phase errors (phase noise) in the additive white Gaussian noise (AWGN) channel. The BEF is the minimum bit error rate (BER) in the presence of phase noise that cannot be reduced by increasing SNR. The phase noise leads to random phase errors on the received signal constellation. Therefore, unlike the BER calculation in the AWGN channel in which the direct distances between the signal points and the bit decision boundaries are deterministic, we compute the angular distances to derive the exact BEF. Simulation results validate the exactness of the proposed expression for different levels of phase noise. Smaller angular distances are equivalent to dominant terms. In order to make the proposed analysis more applicable, a simple formula is then provided by ignoring the non-dominant terms of the exact expression. The evaluation of the simpler formula shows an accurate performance for the smaller values of the phase noise which is the case in most practical scenarios.

Keywords:

MQAM , phase noise , phase error , bit error rate , bit error floor