In recent years the PON research community has focused on future systems targeting 100 Gb/s/ $\lambda$ and beyond, with digital signal processing seen as a key enabling technology. Spectrally efficient 4-level pulse amplitude modulation (PAM4) is seen as a cost-effective solution that exploits the ready availability of cheaper, low-bandwidth devices, and Semiconductor Optical Amplifiers (SOA) are being investigated as receiver preamplifiers to compensate PAM4's high signal-to-noise ratio requirements and meet the demanding 29 dB PON loss budget. However, SOA gain saturation-induced patterning distortion is a concern in the context of PON burst-mode signalling, and the 19.5 dB loud-soft packet dynamic range expected by the most recent ITU-T 50G standards. In this paper we propose a recurrent neural network equalisation technique based on gated recurrent units (GRU-RNN) to not only mitigate SOA patterning affecting loud packet bursts, but to also exploit their remarkable effectiveness at compensating non-linear impairments to unlock the SOA gain saturated regime. Using such an equaliser we demonstrate $ > 28$ dB system dynamic range in 100 Gb/s PAM4 system by using SOA gain compression in conjunction with GRU-RNN equalisation. We find that our proposed GRU-RNN has similar equalisation capabilities as non-linear Volterra, fully connected neural network, and long short-term memory based equalisers, but observe that feedback-based RNN equalisers are more suited to the varying levels of impairment inherent to PON burst-mode signalling due to their low input tap requirements. Recognising issues surrounding hardware implementation of RNNs, we investigate a multi-symbol equalisation scheme to lower the feedback latency requirements of our proposed GRU-RNN. Finally, we compare equaliser complexities and performances according to trainable parameters and real valued multiplication operations, finding that the proposed GRU-RNN equaliser is more efficient than those based on Volterra, fully connected neural networks or long short-term memory units proposed elsewhere.
8 October 2024 in Journal of Optical Communications and Networking
In this paper, we propose an original adaptive neural network equalizer (NNE) algorithm named SkipNet, which is suitable for rapid training on a packet-by-packet basis for burst-mode non-linear equalization in upstream PON transmission. SkipNet uses the simple LMS algorithm and avoids complex neural network training algorithms such as backpropagation and mini-batch training. We demonstrate SkipNet on captured continuous mode 100 Gbit/s PAM4 signals using an SOA preamplifier to achieve the challenging 29 dB PON optical loss budget. The adaptive SkipNet equalizer is shown to overcome combinations of severe SOA patterning effects and fiber dispersion impairments to achieve >29dB dynamic range back-to-back and >22.9dB dynamic range for up to 81.6 ps/nm accumulated dispersion. It can adapt in as little as 250 training symbols to each impairment scenario, which is equivalent to existing FFE/DFE solutions, while matching the non-linear performance of previously proposed static NNE solutions. To the best of our knowledge, SkipNet is the first ever adaptive NNE framework that can realistically be trained and adapted on a packet-by-packet basis and within strict PON packet preamble lengths.
Research on passive optical networks (PONs) capable of reaching speeds of 100 Gb/s per wavelength and beyond has become critical in recent years. One cost-effective solution is 4-level pulse amplitude modulation (PAM4), which utilizes affordable, low-bandwidth devices. The use of semiconductor optical amplifiers (SOAs) as receiver preamplifiers is being investigated to compensate the high PAM4 signal to noise ratio requirement and address PON loss budgets of up to 29 dB. The most recent ITU-T 50G standard specifies an expected loud-soft packet dynamic range of 20.5 dB, currently managed by burst mode trans-impedance amplifiers (BM-TIAs). However, conventional BM-TIAs face limitations in dynamic range and bandwidth when processing high-speed burst-mode upstream signals. To overcome these challenges, we propose burst power equalization in the optical domain to simplify the burst-mode receiver’s design. This paper introduces two novel SOA-based equalization techniques referred to here as: (A) variable bias and (B) control light injection. These methods not only equalize the burst packets but mitigate nonlinear SOA patterning effects in high-power bursts. Our experimental results demonstrate that these SOA-based equalizers extend the dynamic range to 24 dB in a 100 Gb/s PAM4 system, maintaining bit error rates below the soft-decision forward error correction threshold. These findings show that SOA-based equalization reduces the need for high-gain BM-TIAs, thereby preventing any bandwidth limitations associated with the TIA. Furthermore, we show that complex digital signal processing is unnecessary to achieve high dynamic range using an SOA preamplifier, and linear equalization is sufficient to compensate low-cost 25G class receiver bandwidth limitations and relevant fiber dispersion effects. Our approach reduces system complexity while enhancing performance, and therefore represents a cost-effective and scalable solution, positioning SOA-based equalization as a critical innovation for future-proofing high-speed optical networks.
We propose a novel equalization scheme for 100Gb/s/λ PAM4 PON based on Gated Recurrent Neural Network to increase SOA preamplifier input power dynamic range tolerance to 30 dB below hard-decision FEC BER limit of 3.8×10−3.