Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
1535534 | Optics Communications | 2013 | 7 Pages |
We present a comparative study on three carrier phase estimation algorithms, including a one-tap normalized least mean square (NLMS) method, a block-average method, and a Viterbi–Viterbi method in the n-level phase shift keying coherent transmission systems considering the equalization enhanced phase noise (EEPN). In these carrier phase estimation methods, the theoretical bit-error-rate floors based on traditional leading-order Taylor expansion are compared to the practical simulation results, and the tolerable total effective linewidths (involving the transmitter, the local oscillator lasers and the EEPN) for a fixed bit-error-rate floor are evaluated with different block sizes, when the fiber nonlinearities are neglected. The complexity of the three carrier phase estimation methods is also discussed. We find that the carrier phase estimation methods in practical systems should be analyzed based on the simulation results rather than the traditional theoretical predictions, when large EEPN is involved. The one-tap NLMS method can always show an acceptable behavior, while the step size is complicated to optimize. The block-average method is efficient to implement, but it behaves unsatisfactorily when using a large block size. The Viterbi–Viterbi method can show a small improvement compared to the block-average method, while it requires more computational complexity.