The rapid development of data center networks necessitates intensive research into high-speed and high-capacity data center optical transmission systems. Single carrier 400 or 600 Gbit/s transmission will gradually become the mainstream transmission rate in the next generation of communication network. While the application of 400 Gbit/s transmission is well-studied in data center networks, detailed research on 600 Gbit/s transmission systems is less common.

This study is based on a single carrier 600 Gbit/s next-generation data center Elastic Optical Network (EON) transmission system, and conducts detailed theoretical analysis and experimental research on factors affecting transmission distance and spectrum utilization efficiency.

The analysis shows that the maximum transmission distance is determined by the single carrier's peak input optical power, while the spectrum utilization efficiency is related to the bandwidth of the transmission channel. Experimental investigation is conducted on a single carrier 600 Gbit/s EON transmission system. By comparing the relationships between different fiber input optical powers, the system’s Q factor and the pre correction bit error rate, as well as the relationships between 3 dB channel filtering bandwidth and system’s Q factor for different channel numbers, it is confirmed that the system’s transmission distance and spectrum utilization efficiency are related to the optimal fiber input optical power and filtering bandwidth, respectively. Furthermore, the experiment shows that the optimal single-wave input optical power and optimal filtering bandwidth of the 600 Gbit/s transmission system are + 4 dBm and 77 GHz, respectively. Under these settings, the system achieves the longest transmission distance and the highest spectral efficiency. Under these experimental conditions, the 600 Gbit/s transmission system achieves long-term stable operation without error for 48 hours, indicating that this fiber input power and bandwidth can effectively extend and improve the transmission distance and spectrum utilization of the 600 Gbit/s communication system.

The 600 Gbit/s EON system has an optimal input power and filter bandwidth that maximizes bandwidth utilization and transmission distance without significant fiber nonlinear effects or channel interference. The findings have significant values for the engineering construction of 600 Gbit/s transmission systems.

With the advent of new internet services and the broader application of Passive Optical Network (PON) in the industry, PON, as the core technology of the 5th Generation Fixed Networks (F5G), needs to evolve and upgrade continuously to meet the needs of the future network. The latency of PON networks is a key network performance indicator that urgently needs improvement. This article conducts in-depth research on the low latency technologies in the PON networks.

This article first summarizes the challenges faced by current PON networks in terms of low latency. It identifies two main reasons for high latency: the inherent uplink delay introduced by the Optical Network Unit (ONU) bandwidth allocation mechanism and the random uplink delay introduced by the ONU registration/ranging mechanism. To address these issues, this article proposes two ultra-low latency solutions based on single frame multi burst technology and an independent registration channel, followed by experimental verification and analysis.

Analysis shows that the single-frame multi-burst approach optimally balances latency and bandwidth utilization in a quarter scheduling cycle, albeit at the expense of some bandwidth. This method is ideal for latency-sensitive, bandwidth-tolerant services. The independent registration channel solution, by dedicating an extra wavelength for ONU registration/ranging, eliminates the extra delay and jitter from silent periods in standard service channels.

The ultra-low latency solution for PON discussed in this article can effectively reduce network latency, and help promote the large-scale commercialization of emerging applications such as Virtual Reality (VR) and accelerate the growth of industrial internet.

The commonly used max-log-map algorithm reduces the complexity of soft information computation compared to the log-map algorithm. However, it still consumes significant resources for high-order Orthogonal Amplitude Modulation (QAM), such as 8QAM, 16QAM Probability Shaping (PS), and 64QAM PS.

8QAM adopts the method of constellation region division, and the soft information is represented by the distance information from the received symbol to the perpendicular between the nearest bit 0 and bit 1. The distance information is simplified and calculated using angle rotation and regional symmetry. For 16QAM PS and 64QAM PS with non Maxwell Boltzmann (MB) distribution, the central boundary between the edges of the bit 1 region and the central boundary between bit 0 on both sides of the bit 1 region no longer coincide. Region merging approximation is used to handle the region ownership between the two boundaries, and the max-log-map expression is factorized to simplify the distance difference to calculate the soft information. The soft information of 16QAM PS and 64QAM PS based on MB distribution can be obtained by simplifying the soft information expression of non-MB distribution.

The simplification reduced the multiplication and addition/subtraction operations in 8QAM soft information calculation from 48 and 75 times to 12 and 16 times respectively. with a degradation of only about 0.05 dB. For MB-distributed 16QAM PS, operations reduce from 192 multiplications and 260 additions/subtractions to 2 and 4, respectively, also with a degradation of only about 0.05 dB. The reduction is even greater for 64QAM PS, decreasing from 1 152 multiplications and 1 542 additions/subtractions to 3 and 6, respectively.

This article proposes a soft information computation method suitable for 8QAM and MB-distributed 16QAM PS, MB-distributed 64QAM PS, non MB-distributed16QAM PS, and non MB-distributed 64QAM PS. When probabilities align with the MB distribution, the non MB methods can transform into the MB methods. When the shaping factor is 0, the expression based on the MB distribution can be converted into a uniformly distributed soft information calculation formula. The soft information calculation for non-MB distribution, MB distribution, and uniform distribution can be uniformly designed on the same circuit, improving circuit reuse rate and reducing the hardware resource consumption.

Optical fiber dispersion, nonlinear impairments, and bandwidth limitation in high-speed passive optical access networks result in significant power budget loss with traditional intensity modulation and direct detection schemes, failing to meet the requirements of high-speed passive optical access networks.

In order to enhance the rate and performance of the intensity modulation and direct detection optical access system, this study explores the channel equalization method of VDFE-RLS based on Recursive Least Squares (RLS) algorithm, building upon the Volterra Decision Feedback Equalizer (VDFE), This equalizer uses RLS algorithm to update its tap coefficients. It includes a three-order Volterra series, allowing it to compensate for both linear and nonlinear impairments. VDFE-RLS is applied to the downlink optical access system with a single wavelength of 200 Gbit/s based on O-band intensity modulation and direct detection scheme after 20 km transmission.

The experimental results show that the RLS outperforms the conventional Least Mean Square (LMS) algorithm in the equalization process. Moreover, VDFE-RLS achieves a power budget greater than 29 dB. When the length of the equalizer is the same, VDFE-RLS can increase the power budget by 2.2 dB compared with the conventional Volterra Feed Forward Equalization (VFFE)-RLS equalizer. When the length of VDFE-RLS equalizer is half that of VFFE-RLS, the VDFE-RLS can increase the power budget by 0.5 dB compared with the VFFE-RLS equalizer.

Compared with other traditional methods, the proposed method can shorten the length of the equalizer and increase the power budget of the system, ultimately restoring signals with better performance.

In recent years, data communication traffic has experienced explosive growth. To meet the demands for high-speed, high-capacity data transmission and the diverse network application scenarios, hybrid network beyond 100 Gbit/s (B100 Gbit/s) using Dense Wavelength Division Multiplexing (DWDM) has increasingly been recognized as an effective solution. This paper analyzes the requirements, key technologies, and practical case studies of such networks, providing technical support and guidance for building high-capacity and efficient communication networks.

This paper first outlines the requirements for developing B100 Gbit/s DWDM hybrid networks, including network capacity expansion and support for complex network architectures. Next, it details the key technologies for these networks, including constellation shaping, spectrum shaping, and flexible grid technologies. To support bitrate design in hybrid networking, a method for calculating Optical Signal-to-Noise Ratio (OSNR) in cascaded Erbium-Doped Fiber Amplifier (EDFA) communication systems is presented, using parameters such as channel configuration information, transmitted signal optical power, and EDFA gain and noise parameters, to calculate the output OSNR for each wavelength across the link. Finally, by integrating a foreign network case study and based on actual OSNR evaluation, a rational hybrid rate network design is performed, demonstrating the application effectiveness of B100 Gbit/s DWDM hybrid networking in engineering projects.

Implementing B100 Gbit/s DWDM hybrid network, after flexibly configuring transmission rates and bandwidths based on OSNR evaluations, achieves hybrid rate networks deployment at 200, 600 and 800 Gbit/s. This approach fulfills the high-capacity requirements of core sites while accommodating the long-distance, extensive span requirements of edge sites. Furthermore, network upgrades and expansion are smoothly accomplished within a three-year period.

Practice demonstrates that B100 Gbit/s DWDM networking effectively enhances network capacity, flexibility, and spectrum resource utilization. It also provides room for the continuous network evolution, playing a crucial role in advancing the development of high-capacity optical transmission networks.

In order to meet the growing demand for ultra-gigabit, from gigabit to 10-gigabit, from The 5th Generation Fixed networks (F5G) to F5G-Advanced (F5.5G), 50 Gbit/s Passive Optical Network (PON) is considered to be an important part of F5.5G. This paper conducts a thorough study of 50 Gbit/s PON technology, reflecting the latest advancements in access networks.

This paper initially addresses the difficulties in achieving high sensitivity in 50 Gbit/s PON, and proposes a solution. 50 Gbit/s Non-Return-to-Zero (NRZ) signals pass through Avalanche Photodiode (APD) detectors to form macroscopic currents that can be detected under strong electric fields. These currents are then amplified by a Transresistance Amplifier (TIA) and converted into voltage outputs. After balancing, the Feed Forward Equalizer (FFE) and Decision Feedback Equalizer (DFE) of optical Digital Signal Processing (oDSP) chip are used to compensate the trailing phenomenon of pulse signal, and the influence of intersymbol interference is minimized by DFE. Then the paper analyses the key technologies such as APD, TIA and oDSP, comparing the reception performance between 25 Gbit/s and 50 Gbit/s APDs.

Experimental test results show that the 25 Gbit/s APD maintains error-free signal reception at -8.48 dBm over 4 minutes, and reaches -26.61 dBm at a Bit Error Rate(BER) of 2.78e-2. The 50 Gbit/s APD exhibits no bit errors at -8.97 dBm and achieves -27.05 dBm at the same BER, with the second 50 Gbit/s APD set demonstrating similar results.

The 50 Gbit/s APD has superior receiving sensitivity and performance, making it highly suitable for 50 Gbit/s PON optical modules to help achieve high receiver sensitivity. The paper suggests a feasible approach for cost reduction in future implementations, focusing on the balance technology and APD.