Turbo Equalization Techniques Toward Robust PDM 16-QAM Optical Fiber Transmission

Abstract

Abstract-In this paper, we show numerically and experimentally that turbo equalization (TE) is an efficient technique to mitigate performance degradations stemming from optical fiber propagation effects in both optical fiber dispersion managed and unmanaged coherent detection links. The effectiveness of the proposed solution can be appreciated in both linear and nonlinear regimes for either scenario. We report on a system employing a polarization division multiplexing (PDM) 16-quadrature amplitude modulation (QAM) format for which we accomplish an increment in tolerance to link input power of up to 3 dB that represents a substantial improvement margin. The best bit error rate (BER) performances will therefore be guaranteed in a larger window, 6 dB, of link input power thanks to the implemented TE scheme. Moreover, our proposed approach is also proven to effectively mitigate interchannel impairments from surrounding amplitude shift-keying interfering channels in a dispersion managed link achieving also in this case an increment in power tolerance of 3 dB. Furthermore, in terms of BER performances, our proposed TE approach guarantees a gain of about a half order of magnitude at the best operational point. As TE can be included in the current coherent detection transceiver technologies and complement other equalization techniques, it has prospects for application in next-generation high-capacity and long-reach optical transmission links.

Counteracting 16-QAM Optical Fiber Transmission Impairments With Iterative Turbo Equalization

“…The low complexity of the solution proposed, thanks to the presence of a simple convolutional code, and its ability to accomplish for each new iteration a more accurate estimation of the unknown channel, make it a promising solution for high speed, high capacity optical networking. We demonstrate that in a PDM 16 QAM transmission over 251 and 497 km of standard single mode fibre (SSMF) link the use of a Turbo Equalization algorithm improves the bit error rate (BER) performance by more than one order of magnitude, which overcomes previous results reported for compensated link transmission [4]. Moreover we demonstrate that further improvement can be obtained employing additional iterations of the Gray mapping based soft-input soft-output (SISO) TE structure presented.…”

Optical Nonlinear Phase Noise Compensation for <inline-formula><tex-math>$9\times 32$</tex-math> </inline-formula>-Gbaud PolDM-16 QAM Transmission Using a Code-Aided Expectation-Maximization Algorithm

“…However, even with small channel memory, the use of high-order modulation causes the trellis complexity associated with the BCJR algorithm to be very high. A normalized least mean square algorithm was used in [15]- [17]. However, the results were presented based on only a relatively low rate (rate 1/2) convolutional code, which achieved a reported 2 dB launch-power tolerance improvement, measured pre-FEC.…”