Ramon Gutierrez Castrejon

We present a comprehensive and computationally efficient analytical framework to estimate the performance of linear dual-polarization coherent optical systems employing finite-length multiple input multiple output (MIMO) equalizers. By adopting a formulation based on Block-Toeplitz matrices, our model extends prior art by uniquely capturing the complex interplay between optical filtering, chromatic dispersion (CD), distributed polarization impairments, namely polarization mode dispersion (PMD) and polarization dependent loss (PDL), distributed and colored noise, and fractional sampling regimes (e.g., 3/2, 4/3 samples-per-symbol (SpS)) in coherent optical transmission systems. Unlike traditional asymptotic models, we explicitly quantify the impact of limited equalizer memory and colored noise statistics on the post-equalization signal-to-noise-ratio (SNR), which can directly lead to bit error ratio (BER) degradation. Extensive validation against time-domain decision-directed (DD)-least mean square (LMS) simulations, using probabilistic shaping (PS)-16-QAM format, demonstrates the model’s accuracy, showing an average estimation error up to 0.3 dB even under severe impairment scenarios with an SNR penalty ≲15 dB. Furthermore, we leverage the model to highlight the advantages of fractional sampling, showing that 3/2 SpS and 4/3 SpS architectures yield superior performance compared to standard 2 SpS solutions given the same equalizer complexity (i.e., same number of taps) once converging. Notably, this framework offers a dramatic reduction in computational time, by orders of magnitude, compared to traditional time-domain simulations, making it a pivotal tool for the design and optimization of future reconfigurable optical networks.

We demonstrate a method for mitigating the effects of laser phase noise on the detection and localization of vibrations in a bidirectional coherent optical fiber transmission system. By introducing a relative delay between the two frequency tones generated at the transmitter, we retrieve an estimate of the transmitter laser phase noise at the coherent receiver. We demonstrate 200 Gbps, 16 QAM error-free transmission over 80 km while simultaneously detecting and localizing a 3 kHz vibration with 24 rad peak-to-peak phase swing. Compared to conventional methods, our method results in a reduction of localization error by 58%, from 102 m to 43 m, and a reduction in standard deviation of estimated locations by 81%, from 355 m to 67 m.

Este artículo presenta el desarrollo de una interfaz gráfica de código abierto como un recurso de apoyo para que los usuarios del proyecto srsRAN, de SRS (Software Radio Systems), instalen, configuren y visualicen las métricas de desempeño de redes celulares 5G privadas, usando la tecnología SDR (Software Defined Radio) del fabricante Ettus Research (radios USRP). La contribución de este desarrollo de software reside en que posibilita un procedimiento más intuitivo y rápido, a través de un entorno más amigable, para lograr estos propósitos de la red celular 5G, frente al método tradicional de comandos por terminal en el SO Linux. Como parte del trabajo, se exploró el potencial didáctico del software con un grupo de estudiantes de Ingeniería en Telecomunicaciones, mediante su utilización en una práctica de laboratorio, logrando resultados de valoración positivos.

With the unprecedented growth of internet traffic, driven mainly by the recent surge of artificial intelligence and data-intensive applications, the need to address the demand for higher capacity datacenter interconnects is pressing. This invited article presents a study and experimental demonstration for achieving the future generation of 1.6 Tbps Ethernet standards. More specifically, this work presents an empirical demonstration of 1.6 Tbps at 2 and 10 km using the 4$\lambda$ LAN-WDM grid with cost and power effective uncooled DFB lasers, sub-1V driving voltages, low half-wave voltage and high-bandwidth TFLN modulators and quantum-dot SOAs. At 2 km, 1.33 Tbps was achieved using 210 GBaud PAM4 under the 25% OH SD-FEC threshold, 1.6 Tbps using 200 GBaud PAM6 under the 25% OH SD-FEC threshold, and 160 GBaud PAM8 under the 20% OH SD-FEC threshold. At 10 km, due to larger CD, 1.6 Tbps was achieved with asymmetrical channel loading where the edge-most channel used 130 GBaud PAM8 while the remaining three channels used 170 GBaud PAM8 where combined provide an aggregate net throughput of 1.6 Tbps under the 20% OH SD-FEC threshold. These results are followed by an analysis of the impacts of CD-induced power fading on data transmission at 2, 10, 20, and 40 km for channel rates of 100, 200, and 400 Gbps. This analysis reveals that channel rates of 400 Gbps are implementable at distances up to 10 km using WDM grids wide enough to accomodate for the use of uncooled lasers, but are practically unfeasible at distances larger than 10 km. Instead, for these longer reaches, channel rates of 100 Gbps are favorable due to lower accumulated CD.

Due to the exponential growth of internet traffic, the demand for increased capacity for inter and intra-datacenter interconnects is pushing the adoption of coherent and WDM transceivers in the O-band. Semiconductor optical amplifiers are a popular choice of amplifying technology for datacenter applications, but the choice of active region technology (QD, QW and Bulk) remains unclear for signal amplification. This article presents a qualitative and quantitative comparison of SOA and fiber amplifier characteristics and performance for IM/DD and coherent transmission systems operating at the Tbps capacities required for the next generation of optical transceivers. First, we discuss the advantages of QD for SOAs. Next, we quantify key SOA performance metrics for QD, QW and bulk SOA technologies. The QD SOA demonstrates an input saturation power of +4.1dBm, a small signal gain of 18.2dB, a noise figure of 5.6dB and an $\alpha$ -factor of 0.86-1.2. Finally, a high speed transmission performance analysis is presented for single- wavelength and WDM configurations for both IM/DD and coherent transmission. The QD SOA outperformed all other amplifiers for both IM/DD and coherent. Notably, the QD SOA enabled 432Gbps using 180 GBaud PAM8 under the 25% overhead SD-FEC threshold and was the only amplifier to achieve 1.6 Tbps for coherent transmission under the same FEC threshold at 10km. These results underscore the exceptional advantage of QD for SOAs highlighting its potential for future high-capacity optical networks operating in both IM/DD and coherent across single or multiple wavelengths.

Most current Data Center Interconnects (DCI) use intensity modulation direct detection (IMDD) configurations due to their low complexity and cost. However, significant scaling challenges allow coherent solutions to become contenders in these short reach applications. We present an O-band coherent optical fiber transmission system based on Quantum Dot—Mode Locked Lasers (QD-MLLs) using two independent free-running comb lasers, one each for the carrier and the Local Oscillator (LO). Using a comb-to-comb configuration, we demonstrate a 10 km single mode fiber, O-band, coherent, heterodyne, 12.1 Tbps system operating at 0.47 Tbps/λ using 26 λs. We used fewer comb lines (26 λs), faster symbol rate (56 GBaud) and higher constellation cardinality (32 QAM) relative to the highest capacity C-band systems reported to date. Through design, analysis, and experimentation, we quantify the optimum comb line spacing for this use case. We compare potential configurations for increasing data center interconnect capacities whilst reducing power consumption, complexity, and cost.

As coverage increases or network topology becomes more complex, the optical path loss difference between any two access points in tree-like passive optical networks (PON) is uneven, causing service fairness problems at the physical layer. The service fairness of the far-near user access in 200-Gbps coherent PON with various user cases is discussed in this work. A fairness allocation mechanism based on time division multiplexing (TDM) non-orthogonal multiple access (NOMA) in the time-power domain is proposed theoretically to provide fair transmission between the arbitrary two optical network units (ONUs) with the best and the worst performance within single timeslots. Then, a series of subsequent experiments are performed with different path loss and split ratios (SR) to prove the effectiveness and practicality of the proposed NOMA scheme, totally including 12 subcases. Any sub-experiment proves that by optimizing power allocation, the signal-to-interference noise ratio (SINR) or bit-error rate (BER) on the physical layer can be approximated to the same value. The Jain's fairness index at a fixed and flexible hard-decision forward error correction (HD-FEC) standard, and modified fairness index are used to evaluate service fairness in different cases. This TDM-NOMA scheme may address the service fairness issue in next-generation coherent PON with wide coverage.

With the exponential growth of internet traffic, driven by the increasing number of connected devices and data-intensive applications, there is an urgent need to address the surging demand for higher capacity in datacenter communications. This paper proposes using O-band single-carrier coherent transmission to achieve 1.6 Tbps for intra-datacenter reach (2-10 km), leveraging the advancements in next-generation DACs and the TFLN platform. To support our proposal, we assess experimentally the gain of employing 256 GSa/s interleaved DACs compared to the current state-of-the-art 128 GSa/s DACs. Additionally, we explore the feasibility of utilizing cost-effective DFB lasers in these short-reach systems. Using the 128 GSa/s DAC and DFB lasers, we transmit 120 Gbaud DP-64QAM over 10 km of SSMF under the 20% overhead SD-FEC threshold, featuring a net rate of 1.2 Tbps. Switching to 256 GSa/s DAC, we achieve net 1.6 Tbps transmission over 10 km with 167 Gbaud DP-64QAM below the 25% SD-FEC BER threshold. We observe that the power penalty of using DFB lasers compared to ECLs is less than 1 dB. Furthermore, this study includes a comprehensive analysis of the power consumption envelope for various candidate configurations targeting 1.6 Tbps operation. The comparison reveals the competitiveness of the O-band single-carrier coherent solution, attributed to its simpler architecture and the inherent features of the TFLN platform. The analysis highlights the potential of the proposed solution as a power-efficient and high-performance option for meeting the demanding requirements of 1.6 Tbps Ethernet.

We propose two low-complexity digital signal processing (DSP) techniques to improve the system performance of digital subcarrier multiplexing (DSCM) optical transmission system with a silicon photonics transmitter. We first analyze the impacts of various transmitter parameters on the system performance of the single carrier system versus the DSCM system. We show quantitatively that the DSCM system suffers from a high peak-to-average power ratio (PAPR) and nonlinear transfer functions that cannot be easily compensated for. Furthermore, using a high-driving-voltage silicon photonics modulator exacerbates this penalty at the transmitter. To combat the performance degradation caused by nonideal transmitters, we propose and demonstrate the functionality of an encoding scheme based on the fast Fourier transform (FFT) to decrease the PAPR of the transmitted DSCM signals. Then a simple and effective pre-mapping technique is proposed to compensate for the nonlinearity from the transmitter. After developing the theory of our proposed approach, both DSP blocks are verified with coherent optical transmission simulations and experiments. Using a 64 GBd 4-bit/s/Hz DSCM signal containing 8 subcarriers transmitted over 43.2 km of standard single-mode fiber (SSMF), the FFT encoding achieves a gain of 3.458 dB in terms of link loss, and the pre-mapping achieves a gain of 0.486 dB when compared to the raw DSCM system, at the HD-FEC bit error rate (BER) threshold of 3.8e-3. We also tested the performance of the system when the two techniques are combined. We found that this led to a power budget increase of 4.159 dB at the HD-FEC threshold. Since the total gain is more than the addition of the two gains from each DSP block, there is a gain enhancement effect between the two proposed algorithms that generates extra gain when implemented together. The proposed transmitter algorithms and the overall schematic is favorable to the implementation of DSCM systems when using silicon-photonics-modulator-based transmitters specifically and coherent transmitters generally.