Intensity modulation and direct detection (IM/DD) together with pulse amplitude modulation (PAM) is a cost-efficient solution for intra-datacenter links. However, sophisticated digital signal processing (DSP) is inevitable to reach the demanded high data rates for these short-reach applications. In this paper, we compare different DSP schemes that enable rates of 200 Gb/s/ and beyond. Those schemes include Tomlinson-Harashima precoding, partial response signaling, Viterbi equalization and post-filtering together with Volterra nonlinear equalization. Extensive experimental measurements are shown in which the DSP approaches are compared at gross rates between 180 Gb/s and 300 Gb/s. The experiments are done in C-band for fiber links of up to 2 km length. As modulation formats, PAM4, PAM6 and PAM8 are considered. Net bit rates well over 200 Gb/s are shown for all formats at a range of 1 km single mode fiber using a transmission system consisting of currently available commercial components. The optimum DSP configuration as well as the best modulation format varies along the investigated data rates. The best solution is shown to change from partial response PAM4 for lower rates to PAM6 and PAM8 at high bit rates.

Nonlinear distortion has always been a challenge for optical communication due to the nonlinear transfer characteristics of the fiber itself. The next frontier for optical communication is a second type of nonlinearities, which results from optical and electrical components. They become the dominant nonlinearity for shorter reaches. The highest data rates cannot be achieved without effective compensation. A classical countermeasure is receiver-side equalization of nonlinear impairments and memory effects using Volterra series. However, such Volterra equalizers are architecturally complex and their parametrization can be numerical unstable. This contribution proposes an alternative nonlinear equalizer architecture based on machine learning. Its performance is evaluated experimentally on coherent 88 Gbaud dual polarization 16QAM 600 Gb/s back-to-back measurements. The proposed equalizers outperform Volterra and memory polynomial Volterra equalizers up to 6th orders at a target bit-error rate (BER) of ${10}^{-2}$ by 0.5 dB and 0.8 dB in optical signal-to-noise ratio (OSNR), respectively.

We report an industry leading optical dense wavelength division multiplexing (DWDM) field trial with line rates per channel.1.66 Tb/s using 130 GBaud dualpolarization probabilistic constellation shaping 256-ary quadrature amplitude modulation (DP-PCS256QAM) in a high capacity data center interconnect (DCI) scenario. This research trial was performed on 96.5 km of field-deployed standard single mode G.652 fiber infrastructure of Deutsche Telekom in Germany employing Erbium-doped fiber amplifier (EDFA)-only amplification. A total of 34 channels were transmitted with 150 GHz spacing for a total fiber capacity of 56.51 Tb/s and a spectral efficiency higher than 11bit/s/Hz. In the single-channel transmission scenario 1.71 Tb/s was achieved over the same link. In addition, we successfully demonstrate record net bitrates of 1.88 Tb/s in back-to-back (B2B) using 130 GBaud DP-PCS400QAM.

High speed optical transmission systems suffer from intersymbol interference (ISI) and colored noise induced by nonlinear bandwidth limited optical and electrical components. As a countermeasure, this article investigates deep neural network soft-demappers. In particular, we propose a bidirectional recurrent neural network soft-demapper (BRNN-SD) and benchmarked its performance against a time delay neural network soft-demapper (TDNN-SD) and a reference digital signal processing (DSP) scheme consisting of a Volterra nonlinear equalizer accompanied by a symbol-spaced whitening filter and a BCJR detector. On coherent 92GBd dual polarization (DP)-32QAM back-to-back measurements, the proposed soft-demapper matches the performance of the reference DSP. In 800 Gb/s 96GBd DP-32QAM 32-channel dense wavelength division multiplexing (DWDM) transmission over a 600 km fiber link the proposed approach outperforms the reference DSP.

Intensity modulation and direct detection (IM/DD) with pulse amplitude modulation (PAM) is a cost-effective option for intra- and inter-data center (DC) applications. Due to current demand on DC data traffic, transmission rates beyond 200 Gbit/s/ $\lambda $ are desirable. However, employing commercially available lower bandwidth components introduces severe inter-symbol interference (ISI). In addition, to achieving high-speed data rates, digital signal processing (DSP) schemes like Tomlinson-Harashima precoding (THP) and decision feedback equalizer (DFE) can increase tolerance against ISI. In this letter, we show an experimental comparison of PAM-4, PAM-6 and PAM-8 modulation formats with and without THP on the transmitter-side and Volterra nonlinear equalizer (VNLE) on the receiver-side. We provide also results with and without duobinary (DB) encoding on the transmitter side and DFE on the receiver side. The experiment is conducted using an O-band electro-absorption modulated laser over transmission distances up to 20 km.

In this letter, we report on 400 Gb/s polarization multiplexed 16-state quadrature amplitude modulation (PM-16QAM) transponder variants and flex-grid network upgrade configurations. We address transponder subcarrier granularity, and demonstrate that the performance improvement, from dual-carrier to quad-carrier super-channel configuration, is limited to ~ 1.4 dB (in Q-factor, at power spectral density of 10^-1 mW/GHz), at the cost of doubled hardware requirements. In view of that, we establish the performance improvements, for a dual-carrier 400 Gb/s PM-16QAM transceiver, as a function of increasing forward error correction overhead (FEC-OH) and spectral inversion based super-channel fiber nonlinearity compensation (SNLC-SI). We show that increasing the FEC-OH improves the transmission performance, at the cost of significant power consumption requirements, alternatively, employing SNLC-SI, at a lower FEC-OH, is a more power efficient solution. In particular, for homogeneous and heterogeneous launch power based network configurations, SNLC-SI enables ~ 23% and ~ 45% reach improvements at maximum considered FEC-OH (45%). At a fixed distance, it enables ~ 25% and ~ 50% power savings, respectively, compared with FEC-OH employing linear compensation only.